changed name of GainMap to InvertedGainMap

Changes to be able to run the example notebooks: 

- Invert gain map on setting (multiplication is faster but user supplies
ADU/energy)
- Cast after applying gain map not to loose precision (Important for
int32 clusters)
- "factor" for ClusterFileSink 
- Cluster size available to be able to create the right file sink

.clang-tidy

@@ -0,0 +1,42 @@
+
+---
+Checks:          '*, 
+                    -altera-*,
+                    -android-cloexec-fopen,
+                    -cppcoreguidelines-pro-bounds-array-to-pointer-decay, 
+                    -cppcoreguidelines-pro-bounds-pointer-arithmetic, 
+                    -fuchsia*,
+                    -readability-else-after-return,
+                    -readability-avoid-const-params-in-decls,
+                    -readability-identifier-length,
+                    -cppcoreguidelines-pro-bounds-constant-array-index,
+                    -cppcoreguidelines-pro-type-reinterpret-cast,
+                    -llvm-header-guard,
+                    -modernize-use-nodiscard,
+                    -misc-non-private-member-variables-in-classes,
+                    -readability-static-accessed-through-instance,
+                    -readability-braces-around-statements,
+                    -readability-isolate-declaration,
+                    -readability-implicit-bool-conversion,
+                    -readability-identifier-length,
+                    -readability-identifier-naming,
+                    -hicpp-signed-bitwise,
+                    -hicpp-no-array-decay, 
+                    -hicpp-braces-around-statements,
+                    -google-runtime-references, 
+                    -google-readability-todo, 
+                    -google-readability-braces-around-statements,
+                    -modernize-use-trailing-return-type,
+                    -llvmlibc-*'
+
+HeaderFilterRegex: \.hpp
+FormatStyle:     none
+CheckOptions:
+  - { key: readability-identifier-naming.NamespaceCase,       value: lower_case }
+ # - { key: readability-identifier-naming.FunctionCase,        value: lower_case }
+  - { key: readability-identifier-naming.ClassCase,           value: CamelCase  }
+ # - { key: readability-identifier-naming.MethodCase,          value: CamelCase  }
+ # - { key: readability-identifier-naming.StructCase,          value: CamelCase  }
+ # - { key: readability-identifier-naming.VariableCase,        value: lower_case }
+  - { key: readability-identifier-naming.GlobalConstantCase,  value: UPPER_CASE }
+...

.gitea/workflows/cmake_build.yml

@@ -0,0 +1,58 @@
+name: Build the package using cmake then documentation
+
+on:
+  workflow_dispatch:
+
+    
+
+permissions:
+  contents: read
+  pages: write
+  id-token: write
+
+jobs:
+  build:
+    strategy:
+      fail-fast: false
+      matrix:
+        platform: [ubuntu-latest, ] 
+        python-version: ["3.12", ] 
+
+    runs-on: ${{ matrix.platform }}
+
+
+    defaults:
+      run:
+        shell: "bash -l {0}"
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Setup dev env
+        run: |
+            sudo apt-get update
+            sudo apt-get -y install cmake gcc g++
+
+      - name: Get conda
+        uses: conda-incubator/setup-miniconda@v3
+        with:
+          python-version: ${{ matrix.python-version }}
+          environment-file: etc/dev-env.yml
+          miniforge-version: latest
+          channels: conda-forge
+          conda-remove-defaults: "true"
+
+      - name: Build library 
+        run: |
+            mkdir build 
+            cd build 
+            cmake .. -DAARE_SYSTEM_LIBRARIES=ON -DAARE_DOCS=ON 
+            make -j 2 
+            make docs
+
+
+
+
+
+
+

.gitea/workflows/rh8-native.yml

@@ -0,0 +1,36 @@
+name: Build on RHEL8
+
+on:
+  push:
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+jobs:
+  build:
+    runs-on: "ubuntu-latest"
+    container:
+      image: gitea.psi.ch/images/rhel8-developer-gitea-actions
+    steps:
+      # workaround until actions/checkout@v4 is available for RH8
+      # - uses: actions/checkout@v4
+      - name: Clone repository
+        run: |
+          echo Cloning ${{ github.ref_name }}
+          git clone https://${{secrets.GITHUB_TOKEN}}@gitea.psi.ch/${{ github.repository }}.git --branch=${{ github.ref_name }} .
+
+
+      - name: Install dependencies
+        run: |
+          dnf install -y cmake python3.12 python3.12-devel python3.12-pip
+
+      - name: Build library 
+        run: |
+            mkdir build && cd build
+            cmake .. -DAARE_PYTHON_BINDINGS=ON -DAARE_TESTS=ON -DPython_FIND_VIRTUALENV=FIRST
+            make -j 2 
+
+      - name: C++ unit tests
+        working-directory: ${{gitea.workspace}}/build
+        run: ctest

.gitea/workflows/rh9-native.yml

@@ -0,0 +1,31 @@
+name: Build on RHEL9
+
+on:
+  push:
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+jobs:
+  build:
+    runs-on: "ubuntu-latest"
+    container:
+      image: gitea.psi.ch/images/rhel9-developer-gitea-actions
+    steps:
+      - uses: actions/checkout@v4
+
+
+      - name: Install dependencies
+        run: |
+          dnf install -y cmake python3.12 python3.12-devel python3.12-pip
+
+      - name: Build library 
+        run: |
+            mkdir build && cd build
+            cmake .. -DAARE_PYTHON_BINDINGS=ON -DAARE_TESTS=ON
+            make -j 2 
+
+      - name: C++ unit tests
+        working-directory: ${{gitea.workspace}}/build
+        run: ctest

.github/workflows/build_docs.yml

@@ -5,7 +5,6 @@ on:
   push:
     
 
-
 permissions:
   contents: read
   pages: write
@@ -16,12 +15,11 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        platform: [ubuntu-latest, ] # macos-12, windows-2019]
+        platform: [ubuntu-latest, ] 
         python-version: ["3.12",]
 
     runs-on: ${{ matrix.platform }}
 
-    # The setup-miniconda action needs this to activate miniconda
     defaults:
       run:
         shell: "bash -l {0}"
@@ -30,13 +28,13 @@ jobs:
       - uses: actions/checkout@v4
 
       - name: Get conda
-        uses: conda-incubator/setup-miniconda@v3.0.4
+        uses: conda-incubator/setup-miniconda@v3
         with:
           python-version: ${{ matrix.python-version }}
+          environment-file: etc/dev-env.yml
+          miniforge-version: latest
           channels: conda-forge
-
-      - name: Prepare
-        run: conda install doxygen sphinx=7.1.2 breathe pybind11 sphinx_rtd_theme furo nlohmann_json zeromq fmt numpy
+          conda-remove-defaults: "true"
 
       - name: Build library 
         run: |

.github/workflows/build_wheel.yml

@@ -0,0 +1,64 @@
+name: Build wheel
+
+on:
+    workflow_dispatch:
+    pull_request:
+    push:
+      branches:
+        - main
+    release:
+      types:
+        - published
+  
+
+jobs:
+  build_wheels:
+    name: Build wheels on ${{ matrix.os }}
+    runs-on: ${{ matrix.os }}
+    strategy:
+      matrix:
+        os: [ubuntu-latest,]
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Build wheels
+        run: pipx run cibuildwheel==2.23.0
+
+      - uses: actions/upload-artifact@v4
+        with:
+          name: cibw-wheels-${{ matrix.os }}-${{ strategy.job-index }}
+          path: ./wheelhouse/*.whl
+    
+  build_sdist:
+    name: Build source distribution
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v4
+
+    - name: Build sdist
+      run: pipx run build --sdist
+
+    - uses: actions/upload-artifact@v4
+      with:
+        name: cibw-sdist
+        path: dist/*.tar.gz
+
+  upload_pypi:
+    needs: [build_wheels, build_sdist]
+    runs-on: ubuntu-latest
+    environment: pypi
+    permissions:
+        id-token: write
+    if: github.event_name == 'release' && github.event.action == 'published'
+    # or, alternatively, upload to PyPI on every tag starting with 'v' (remove on: release above to use this)
+    # if: github.event_name == 'push' && startsWith(github.ref, 'refs/tags/v')
+    steps:
+        - uses: actions/download-artifact@v4
+          with:
+            # unpacks all CIBW artifacts into dist/
+            pattern: cibw-*
+            path: dist
+            merge-multiple: true
+
+        - uses: pypa/gh-action-pypi-publish@release/v1

.gitignore

@@ -17,7 +17,8 @@ Testing/
 ctbDict.cpp
 ctbDict.h
 
-
+wheelhouse/
+dist/
 
 *.pyc
 */__pycache__/*

CMakeLists.txt

@@ -1,4 +1,4 @@
-cmake_minimum_required(VERSION 3.14)
+cmake_minimum_required(VERSION 3.15)
 
 project(aare 
     VERSION 1.0.0
@@ -11,6 +11,14 @@ set(CMAKE_CXX_STANDARD 17)
 set(CMAKE_CXX_STANDARD_REQUIRED ON)
 set(CMAKE_CXX_EXTENSIONS OFF)
 
+execute_process(
+        COMMAND git log -1 --format=%h
+        WORKING_DIRECTORY ${CMAKE_CURRENT_LIST_DIR}
+        OUTPUT_VARIABLE GIT_HASH
+        OUTPUT_STRIP_TRAILING_WHITESPACE
+        )
+message(STATUS "Building from git hash: ${GIT_HASH}")
+
 if (${CMAKE_VERSION} VERSION_GREATER "3.24")
     cmake_policy(SET CMP0135 NEW) #Fetch content download timestamp
 endif()
@@ -31,7 +39,7 @@ set(CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/cmake" ${CMAKE_MODULE_PATH})
 
 
 # General options
-option(AARE_PYTHON_BINDINGS "Build python bindings" ON)
+option(AARE_PYTHON_BINDINGS "Build python bindings" OFF)
 option(AARE_TESTS "Build tests" OFF)
 option(AARE_BENCHMARKS "Build benchmarks" OFF)
 option(AARE_EXAMPLES "Build examples" OFF)
@@ -60,6 +68,8 @@ if(AARE_SYSTEM_LIBRARIES)
     set(AARE_FETCH_CATCH OFF CACHE BOOL "Disabled FetchContent for catch2" FORCE)
     set(AARE_FETCH_JSON OFF CACHE BOOL "Disabled FetchContent for nlohmann::json" FORCE)
     set(AARE_FETCH_ZMQ OFF CACHE BOOL "Disabled FetchContent for libzmq" FORCE)
+    # Still fetch lmfit when setting AARE_SYSTEM_LIBRARIES since this is not available
+    # on conda-forge
 endif()
 
 if(AARE_VERBOSE)
@@ -78,15 +88,31 @@ endif()
 set(CMAKE_EXPORT_COMPILE_COMMANDS ON)
 
 if(AARE_FETCH_LMFIT)
-    set(lmfit_patch git apply ${CMAKE_CURRENT_SOURCE_DIR}/patches/lmfit.patch)
-    FetchContent_Declare(
-        lmfit
-        GIT_REPOSITORY https://jugit.fz-juelich.de/mlz/lmfit.git
-        GIT_TAG        main
-        PATCH_COMMAND ${lmfit_patch}
-        UPDATE_DISCONNECTED 1
-        EXCLUDE_FROM_ALL
-    )
+    #TODO! Should we fetch lmfit from the web or inlcude a tar.gz in the repo?
+    set(LMFIT_PATCH_COMMAND git apply ${CMAKE_CURRENT_SOURCE_DIR}/patches/lmfit.patch)
+
+    # For cmake < 3.28 we can't supply EXCLUDE_FROM_ALL to FetchContent_Declare
+    # so we need this workaround
+    if (${CMAKE_VERSION} VERSION_LESS "3.28")
+        FetchContent_Declare(
+            lmfit
+            GIT_REPOSITORY https://jugit.fz-juelich.de/mlz/lmfit.git
+            GIT_TAG        main
+            PATCH_COMMAND ${LMFIT_PATCH_COMMAND}
+            UPDATE_DISCONNECTED 1
+        )
+    else()
+        FetchContent_Declare(
+            lmfit
+            GIT_REPOSITORY https://jugit.fz-juelich.de/mlz/lmfit.git
+            GIT_TAG        main
+            PATCH_COMMAND ${LMFIT_PATCH_COMMAND}
+            UPDATE_DISCONNECTED 1
+            EXCLUDE_FROM_ALL 1
+        )
+    endif()
+    
+
     #Disable what we don't need from lmfit
     set(BUILD_TESTING OFF CACHE BOOL "")
     set(LMFIT_CPPTEST OFF CACHE BOOL "")
@@ -94,12 +120,16 @@ if(AARE_FETCH_LMFIT)
     set(LMFIT_CPPTEST OFF CACHE BOOL "")
     set(BUILD_SHARED_LIBS OFF CACHE BOOL "")
 
+    if (${CMAKE_VERSION} VERSION_LESS "3.28")
+        if(NOT lmfit_POPULATED)
+            FetchContent_Populate(lmfit)
+            add_subdirectory(${lmfit_SOURCE_DIR} ${lmfit_BINARY_DIR} EXCLUDE_FROM_ALL)
+        endif()
+    else()
+        FetchContent_MakeAvailable(lmfit)
+    endif()
     
-    FetchContent_MakeAvailable(lmfit)
     set_property(TARGET lmfit PROPERTY POSITION_INDEPENDENT_CODE ON)
-
-    target_include_directories (lmfit PUBLIC "${libzmq_SOURCE_DIR}/lib")
-    message(STATUS "lmfit include dir: ${lmfit_SOURCE_DIR}/lib")
 else()
     find_package(lmfit REQUIRED)
 endif()
@@ -111,10 +141,13 @@ if(AARE_FETCH_ZMQ)
     if (${CMAKE_VERSION} VERSION_GREATER_EQUAL "3.30")
         cmake_policy(SET CMP0169 OLD)
     endif()
+    set(ZMQ_PATCH_COMMAND git apply ${CMAKE_CURRENT_SOURCE_DIR}/patches/libzmq_cmake_version.patch)
     FetchContent_Declare(
         libzmq
         GIT_REPOSITORY https://github.com/zeromq/libzmq.git
         GIT_TAG        v4.3.4
+        PATCH_COMMAND ${ZMQ_PATCH_COMMAND}
+        UPDATE_DISCONNECTED 1
     )
     # Disable unwanted options from libzmq
     set(BUILD_TESTS OFF CACHE BOOL "Switch off libzmq test build")
@@ -307,6 +340,8 @@ endif()
 
 set(PUBLICHEADERS
     include/aare/ArrayExpr.hpp
+    include/aare/CalculateEta.hpp
+    include/aare/Cluster.hpp
     include/aare/ClusterFinder.hpp
     include/aare/ClusterFile.hpp
     include/aare/CtbRawFile.hpp
@@ -317,8 +352,11 @@ set(PUBLICHEADERS
     include/aare/File.hpp
     include/aare/Fit.hpp
     include/aare/FileInterface.hpp
+    include/aare/FilePtr.hpp
     include/aare/Frame.hpp
+    include/aare/GainMap.hpp
     include/aare/geo_helpers.hpp
+    include/aare/JungfrauDataFile.hpp
     include/aare/NDArray.hpp
     include/aare/NDView.hpp
     include/aare/NumpyFile.hpp
@@ -330,38 +368,38 @@ set(PUBLICHEADERS
     include/aare/RawSubFile.hpp
     include/aare/VarClusterFinder.hpp
     include/aare/utils/task.hpp
-
 )
 
 
 set(SourceFiles
     ${CMAKE_CURRENT_SOURCE_DIR}/src/CtbRawFile.cpp
-    ${CMAKE_CURRENT_SOURCE_DIR}/src/ClusterFile.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/defs.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/Dtype.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/decode.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/Frame.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/File.cpp
+    ${CMAKE_CURRENT_SOURCE_DIR}/src/FilePtr.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/Fit.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/geo_helpers.cpp
+    ${CMAKE_CURRENT_SOURCE_DIR}/src/JungfrauDataFile.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/NumpyFile.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/NumpyHelpers.cpp
+    ${CMAKE_CURRENT_SOURCE_DIR}/src/Interpolator.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/PixelMap.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/RawFile.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/RawSubFile.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/RawMasterFile.cpp
-    ${CMAKE_CURRENT_SOURCE_DIR}/src/utils/task.cpp
-)
 
+    ${CMAKE_CURRENT_SOURCE_DIR}/src/utils/task.cpp
+    ${CMAKE_CURRENT_SOURCE_DIR}/src/utils/ifstream_helpers.cpp
+)
 
 add_library(aare_core STATIC ${SourceFiles})
 target_include_directories(aare_core PUBLIC 
     "$<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/include>" 
-    "$<INSTALL_INTERFACE:${CMAKE_INSTALL_INCLUDEDIR}>"
+    "$<INSTALL_INTERFACE:${CMAKE_INSTALL_INCLUDEDIR}>" 
 )
 
-
-
 target_link_libraries(
     aare_core 
     PUBLIC 
@@ -370,7 +408,8 @@ target_link_libraries(
     ${STD_FS_LIB} # from helpers.cmake
     PRIVATE 
     aare_compiler_flags 
-    lmfit
+    $<BUILD_INTERFACE:lmfit>
+
 )
 
 set_target_properties(aare_core PROPERTIES
@@ -384,7 +423,9 @@ endif()
 
 if(AARE_TESTS)
     set(TestSources
+            ${CMAKE_CURRENT_SOURCE_DIR}/src/algorithm.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/defs.test.cpp
+            ${CMAKE_CURRENT_SOURCE_DIR}/src/decode.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/Dtype.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/Frame.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/geo_helpers.test.cpp
@@ -393,7 +434,12 @@ if(AARE_TESTS)
             ${CMAKE_CURRENT_SOURCE_DIR}/src/NDView.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/ClusterFinder.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/ClusterVector.test.cpp
+            ${CMAKE_CURRENT_SOURCE_DIR}/src/Cluster.test.cpp
+            ${CMAKE_CURRENT_SOURCE_DIR}/src/CalculateEta.test.cpp
+            ${CMAKE_CURRENT_SOURCE_DIR}/src/ClusterFile.test.cpp
+            ${CMAKE_CURRENT_SOURCE_DIR}/src/ClusterFinderMT.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/Pedestal.test.cpp
+            ${CMAKE_CURRENT_SOURCE_DIR}/src/JungfrauDataFile.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/NumpyFile.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/NumpyHelpers.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/RawFile.test.cpp

benchmarks/CMakeLists.txt

@@ -1,11 +1,27 @@
-find_package(benchmark REQUIRED)
 
-add_executable(ndarray_benchmark ndarray_benchmark.cpp)
+include(FetchContent)
 
-target_link_libraries(ndarray_benchmark benchmark::benchmark aare_core aare_compiler_flags)
-# target_link_libraries(tests PRIVATE aare_core aare_compiler_flags)
 
-set_target_properties(ndarray_benchmark PROPERTIES
-        RUNTIME_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}
-        # OUTPUT_NAME run_tests
+FetchContent_Declare(
+    benchmark
+    GIT_REPOSITORY https://github.com/google/benchmark.git
+    GIT_TAG        v1.8.3 # Change to the latest version if needed
+)
+
+# Ensure Google Benchmark is built correctly
+set(BENCHMARK_ENABLE_TESTING OFF CACHE BOOL "" FORCE)
+
+FetchContent_MakeAvailable(benchmark)
+
+add_executable(benchmarks)
+
+target_sources(benchmarks PRIVATE ndarray_benchmark.cpp calculateeta_benchmark.cpp)
+
+# Link Google Benchmark and other necessary libraries
+target_link_libraries(benchmarks PRIVATE benchmark::benchmark aare_core aare_compiler_flags)
+
+# Set output properties
+set_target_properties(benchmarks PROPERTIES
+    RUNTIME_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}
+    OUTPUT_NAME run_benchmarks
 )

benchmarks/calculateeta_benchmark.cpp

@@ -0,0 +1,70 @@
+#include "aare/CalculateEta.hpp"
+#include "aare/ClusterFile.hpp"
+#include <benchmark/benchmark.h>
+
+using namespace aare;
+
+class ClusterFixture : public benchmark::Fixture {
+  public:
+    Cluster<int, 2, 2> cluster_2x2{};
+    Cluster<int, 3, 3> cluster_3x3{};
+
+  private:
+    using benchmark::Fixture::SetUp;
+
+    void SetUp([[maybe_unused]] const benchmark::State &state) override {
+        int temp_data[4] = {1, 2, 3, 1};
+        std::copy(std::begin(temp_data), std::end(temp_data),
+                  std::begin(cluster_2x2.data));
+
+        cluster_2x2.x = 0;
+        cluster_2x2.y = 0;
+
+        int temp_data2[9] = {1, 2, 3, 1, 3, 4, 5, 1, 20};
+        std::copy(std::begin(temp_data2), std::end(temp_data2),
+                  std::begin(cluster_3x3.data));
+
+        cluster_3x3.x = 0;
+        cluster_3x3.y = 0;
+    }
+
+    // void TearDown(::benchmark::State& state) {
+    // }
+};
+
+BENCHMARK_F(ClusterFixture, Calculate2x2Eta)(benchmark::State &st) {
+    for (auto _ : st) {
+        // This code gets timed
+        Eta2 eta = calculate_eta2(cluster_2x2);
+        benchmark::DoNotOptimize(eta);
+    }
+}
+
+// almost takes double the time
+BENCHMARK_F(ClusterFixture,
+            CalculateGeneralEtaFor2x2Cluster)(benchmark::State &st) {
+    for (auto _ : st) {
+        // This code gets timed
+        Eta2 eta = calculate_eta2<int, 2, 2>(cluster_2x2);
+        benchmark::DoNotOptimize(eta);
+    }
+}
+
+BENCHMARK_F(ClusterFixture, Calculate3x3Eta)(benchmark::State &st) {
+    for (auto _ : st) {
+        // This code gets timed
+        Eta2 eta = calculate_eta2(cluster_3x3);
+        benchmark::DoNotOptimize(eta);
+    }
+}
+
+// almost takes double the time
+BENCHMARK_F(ClusterFixture,
+            CalculateGeneralEtaFor3x3Cluster)(benchmark::State &st) {
+    for (auto _ : st) {
+        // This code gets timed
+        Eta2 eta = calculate_eta2<int, 3, 3>(cluster_3x3);
+        benchmark::DoNotOptimize(eta);
+    }
+}
+// BENCHMARK_MAIN();

conda-recipe/meta.yaml

@@ -1,6 +1,10 @@
 package:
   name: aare
-  version: 2025.2.12 #TODO! how to not duplicate this?
+  version: 2025.4.22 #TODO! how to not duplicate this?
+
+
+
+
 
 
 source:
@@ -35,6 +39,7 @@ requirements:
   run:
     - python {{python}}
     - numpy {{ numpy }}
+    - matplotlib
 
 
 test:

docs/src/JungfrauDataFile.rst

@@ -0,0 +1,25 @@
+JungfrauDataFile
+==================
+
+JungfrauDataFile is a class to read the .dat files that are produced by Aldo's receiver. 
+It is mostly used for calibration. 
+
+The structure of the file is:
+
+* JungfrauDataHeader
+* Binary data (256x256, 256x1024 or 512x1024)
+* JungfrauDataHeader
+* ...
+
+There is no metadata indicating number of frames or the size of the image, but this
+will be infered by this reader. 
+
+.. doxygenstruct:: aare::JungfrauDataHeader
+   :members:
+   :undoc-members:
+   :private-members:
+
+.. doxygenclass:: aare::JungfrauDataFile
+   :members:
+   :undoc-members:
+   :private-members:

docs/src/Tests.rst

@@ -0,0 +1,47 @@
+****************
+Tests
+****************
+
+We test the code both from the C++ and Python API. By default only tests that does not require image data is run. 
+
+C++
+~~~~~~~~~~~~~~~~~~
+
+.. code-block:: bash
+
+    mkdir build
+    cd build
+    cmake .. -DAARE_TESTS=ON
+    make -j 4
+
+    export AARE_TEST_DATA=/path/to/test/data
+    ./run_test [.files] #or using ctest, [.files] is the option to include tests needing data
+
+
+
+Python
+~~~~~~~~~~~~~~~~~~
+
+.. code-block:: bash
+
+    #From the root dir of the library
+    python -m pytest python/tests --files # passing --files will run the tests needing data
+
+
+
+Getting the test data
+~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. attention ::
+
+    The tests needing the test data are not run by default. To make the data available, you need to set the environment variable
+    AARE_TEST_DATA to the path of the test data directory. Then pass either [.files] for the C++ tests or --files for Python
+
+The image files needed for the test are large and are not included in the repository. They are stored
+using GIT LFS in a separate repository. To get the test data, you need to clone the repository.
+To do this, you need to have GIT LFS installed. You can find instructions on how to install it here: https://git-lfs.github.com/
+Once you have GIT LFS installed, you can clone the repository like any normal repo using:
+
+.. code-block:: bash
+    
+    git clone https://gitea.psi.ch/detectors/aare-test-data.git

docs/src/algorithm.rst

@@ -0,0 +1,5 @@
+algorithm
+=============
+
+.. doxygenfile:: algorithm.hpp
+

docs/src/index.rst

@@ -20,9 +20,6 @@ AARE
     Requirements
     Consume
 
-
-
-
 .. toctree::
     :caption: Python API
     :maxdepth: 1
@@ -31,6 +28,7 @@ AARE
     pyCtbRawFile
     pyClusterFile
     pyClusterVector
+    pyJungfrauDataFile
     pyRawFile
     pyRawMasterFile
     pyVarClusterFinder
@@ -42,6 +40,7 @@ AARE
     :caption: C++ API
     :maxdepth: 1
     
+    algorithm
     NDArray
     NDView
     Frame
@@ -51,6 +50,7 @@ AARE
     ClusterFinderMT
     ClusterFile
     ClusterVector
+    JungfrauDataFile
     Pedestal
     RawFile
     RawSubFile
@@ -59,4 +59,8 @@ AARE
 
 
 
-
+.. toctree::
+    :caption: Developer
+    :maxdepth: 3
+    
+    Tests

docs/src/pyJungfrauDataFile.rst

@@ -0,0 +1,10 @@
+JungfrauDataFile
+===================
+
+.. py:currentmodule:: aare
+
+.. autoclass:: JungfrauDataFile
+    :members:
+    :undoc-members:
+    :show-inheritance:
+    :inherited-members:

etc/dev-env.yml

@@ -0,0 +1,15 @@
+name: dev-environment
+channels:
+  - conda-forge
+dependencies:
+  - anaconda-client
+  - doxygen 
+  - sphinx=7.1.2 
+  - breathe 
+  - pybind11 
+  - sphinx_rtd_theme 
+  - furo 
+  - nlohmann_json 
+  - zeromq 
+  - fmt 
+  - numpy

include/aare/CalculateEta.hpp

@@ -0,0 +1,170 @@
+#pragma once
+
+#include "aare/Cluster.hpp"
+#include "aare/ClusterVector.hpp"
+#include "aare/NDArray.hpp"
+
+namespace aare {
+
+enum class corner : int {
+    cBottomLeft = 0,
+    cBottomRight = 1,
+    cTopLeft = 2,
+    cTopRight = 3
+};
+
+enum class pixel : int {
+    pBottomLeft = 0,
+    pBottom = 1,
+    pBottomRight = 2,
+    pLeft = 3,
+    pCenter = 4,
+    pRight = 5,
+    pTopLeft = 6,
+    pTop = 7,
+    pTopRight = 8
+};
+
+template <typename T> struct Eta2 {
+    double x;
+    double y;
+    int c;
+    T sum;
+};
+
+/**
+ * @brief Calculate the eta2 values for all clusters in a Clustervector
+ */
+template <typename ClusterType,
+          typename = std::enable_if_t<is_cluster_v<ClusterType>>>
+NDArray<double, 2> calculate_eta2(const ClusterVector<ClusterType> &clusters) {
+    NDArray<double, 2> eta2({static_cast<int64_t>(clusters.size()), 2});
+
+    for (size_t i = 0; i < clusters.size(); i++) {
+        auto e = calculate_eta2(clusters[i]);
+        eta2(i, 0) = e.x;
+        eta2(i, 1) = e.y;
+    }
+
+    return eta2;
+}
+
+/**
+ * @brief Calculate the eta2 values for a generic sized cluster and return them
+ * in a Eta2 struct containing etay, etax and the index of the respective 2x2
+ * subcluster.
+ */
+template <typename T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType>
+Eta2<T>
+calculate_eta2(const Cluster<T, ClusterSizeX, ClusterSizeY, CoordType> &cl) {
+    Eta2<T> eta{};
+
+    auto max_sum = cl.max_sum_2x2();
+    eta.sum = max_sum.first;
+    auto c = max_sum.second;
+
+    size_t cluster_center_index =
+        (ClusterSizeX / 2) + (ClusterSizeY / 2) * ClusterSizeX;
+
+    size_t index_bottom_left_max_2x2_subcluster =
+        (int(c / (ClusterSizeX - 1))) * ClusterSizeX + c % (ClusterSizeX - 1);
+
+    // check that cluster center is in max subcluster
+    if (cluster_center_index != index_bottom_left_max_2x2_subcluster &&
+        cluster_center_index != index_bottom_left_max_2x2_subcluster + 1 &&
+        cluster_center_index !=
+            index_bottom_left_max_2x2_subcluster + ClusterSizeX &&
+        cluster_center_index !=
+            index_bottom_left_max_2x2_subcluster + ClusterSizeX + 1)
+        throw std::runtime_error("Photon center is not in max 2x2_subcluster");
+
+    if ((cluster_center_index - index_bottom_left_max_2x2_subcluster) %
+            ClusterSizeX ==
+        0) {
+        if ((cl.data[cluster_center_index + 1] +
+             cl.data[cluster_center_index]) != 0)
+
+            eta.x = static_cast<double>(cl.data[cluster_center_index + 1]) /
+                    static_cast<double>((cl.data[cluster_center_index + 1] +
+                                         cl.data[cluster_center_index]));
+    } else {
+        if ((cl.data[cluster_center_index] +
+             cl.data[cluster_center_index - 1]) != 0)
+
+            eta.x = static_cast<double>(cl.data[cluster_center_index]) /
+                    static_cast<double>((cl.data[cluster_center_index - 1] +
+                                         cl.data[cluster_center_index]));
+    }
+    if ((cluster_center_index - index_bottom_left_max_2x2_subcluster) /
+            ClusterSizeX <
+        1) {
+        assert(cluster_center_index + ClusterSizeX <
+               ClusterSizeX * ClusterSizeY); // suppress warning
+        if ((cl.data[cluster_center_index] +
+             cl.data[cluster_center_index + ClusterSizeX]) != 0)
+            eta.y = static_cast<double>(
+                        cl.data[cluster_center_index + ClusterSizeX]) /
+                    static_cast<double>(
+                        (cl.data[cluster_center_index] +
+                         cl.data[cluster_center_index + ClusterSizeX]));
+    } else {
+        if ((cl.data[cluster_center_index] +
+             cl.data[cluster_center_index - ClusterSizeX]) != 0)
+            eta.y = static_cast<double>(cl.data[cluster_center_index]) /
+                    static_cast<double>(
+                        (cl.data[cluster_center_index] +
+                         cl.data[cluster_center_index - ClusterSizeX]));
+    }
+
+    eta.c = c; // TODO only supported for 2x2 and 3x3 clusters -> at least no
+               // underyling enum class
+    return eta;
+}
+
+// TODO! Look up eta2 calculation - photon center should be top right corner
+template <typename T>
+Eta2<T> calculate_eta2(const Cluster<T, 2, 2, int16_t> &cl) {
+    Eta2<T> eta{};
+
+    if ((cl.data[0] + cl.data[1]) != 0)
+        eta.x = static_cast<double>(cl.data[1]) / (cl.data[0] + cl.data[1]);
+    if ((cl.data[0] + cl.data[2]) != 0)
+        eta.y = static_cast<double>(cl.data[2]) / (cl.data[0] + cl.data[2]);
+    eta.sum = cl.sum();
+    eta.c = static_cast<int>(corner::cBottomLeft); // TODO! This is not correct,
+                                                   // but need to put something
+    return eta;
+}
+
+// calculates Eta3 for 3x3 cluster based on code from analyze_cluster
+// TODO only supported for 3x3 Clusters
+template <typename T> Eta2<T> calculate_eta3(const Cluster<T, 3, 3> &cl) {
+
+    Eta2<T> eta{};
+
+    T sum = 0;
+
+    std::for_each(std::begin(cl.data), std::end(cl.data),
+                  [&sum](T x) { sum += x; });
+
+    eta.sum = sum;
+
+    eta.c = corner::cBottomLeft;
+
+    if ((cl.data[3] + cl.data[4] + cl.data[5]) != 0)
+
+        eta.x = static_cast<double>(-cl.data[3] + cl.data[3 + 2]) /
+
+                (cl.data[3] + cl.data[4] + cl.data[5]);
+
+    if ((cl.data[1] + cl.data[4] + cl.data[7]) != 0)
+
+        eta.y = static_cast<double>(-cl.data[1] + cl.data[2 * 3 + 1]) /
+
+                (cl.data[1] + cl.data[4] + cl.data[7]);
+
+    return eta;
+}
+
+} // namespace aare

include/aare/Cluster.hpp

@@ -0,0 +1,86 @@
+
+/************************************************
+ * @file Cluster.hpp
+ * @short definition of cluster, where CoordType (x,y) give
+ * the cluster center coordinates and data the actual cluster data
+ * cluster size is given as template parameters
+ ***********************************************/
+
+#pragma once
+
+#include <algorithm>
+#include <array>
+#include <cstdint>
+#include <numeric>
+#include <type_traits>
+
+namespace aare {
+
+// requires clause c++20 maybe update
+template <typename T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType = int16_t>
+struct Cluster {
+
+    static_assert(std::is_arithmetic_v<T>, "T needs to be an arithmetic type");
+    static_assert(std::is_integral_v<CoordType>,
+                  "CoordType needs to be an integral type");
+    static_assert(ClusterSizeX > 0 && ClusterSizeY > 0,
+                  "Cluster sizes must be bigger than zero");
+
+    CoordType x;
+    CoordType y;
+    std::array<T, ClusterSizeX * ClusterSizeY> data;
+
+    static constexpr uint8_t cluster_size_x = ClusterSizeX;
+    static constexpr uint8_t cluster_size_y = ClusterSizeY;
+    using value_type = T;
+    using coord_type = CoordType;
+
+    T sum() const { return std::accumulate(data.begin(), data.end(), T{}); }
+
+    std::pair<T, int> max_sum_2x2() const {
+
+        if constexpr (cluster_size_x == 3 && cluster_size_y == 3) {
+            std::array<T, 4> sum_2x2_subclusters;
+            sum_2x2_subclusters[0] = data[0] + data[1] + data[3] + data[4];
+            sum_2x2_subclusters[1] = data[1] + data[2] + data[4] + data[5];
+            sum_2x2_subclusters[2] = data[3] + data[4] + data[6] + data[7];
+            sum_2x2_subclusters[3] = data[4] + data[5] + data[7] + data[8];
+            int index = std::max_element(sum_2x2_subclusters.begin(),
+                                         sum_2x2_subclusters.end()) -
+                        sum_2x2_subclusters.begin();
+            return std::make_pair(sum_2x2_subclusters[index], index);
+        } else if constexpr (cluster_size_x == 2 && cluster_size_y == 2) {
+            return std::make_pair(data[0] + data[1] + data[2] + data[3], 0);
+        } else {
+            constexpr size_t num_2x2_subclusters =
+                (ClusterSizeX - 1) * (ClusterSizeY - 1);
+
+            std::array<T, num_2x2_subclusters> sum_2x2_subcluster;
+            for (size_t i = 0; i < ClusterSizeY - 1; ++i) {
+                for (size_t j = 0; j < ClusterSizeX - 1; ++j)
+                    sum_2x2_subcluster[i * (ClusterSizeX - 1) + j] =
+                        data[i * ClusterSizeX + j] +
+                        data[i * ClusterSizeX + j + 1] +
+                        data[(i + 1) * ClusterSizeX + j] +
+                        data[(i + 1) * ClusterSizeX + j + 1];
+            }
+
+            int index = std::max_element(sum_2x2_subcluster.begin(),
+                                         sum_2x2_subcluster.end()) -
+                        sum_2x2_subcluster.begin();
+            return std::make_pair(sum_2x2_subcluster[index], index);
+        }
+    }
+};
+
+// Type Traits for is_cluster_type
+template <typename T>
+struct is_cluster : std::false_type {}; // Default case: Not a Cluster
+
+template <typename T, uint8_t X, uint8_t Y, typename CoordType>
+struct is_cluster<Cluster<T, X, Y, CoordType>> : std::true_type {}; // Cluster
+
+template <typename T> constexpr bool is_cluster_v = is_cluster<T>::value;
+
+} // namespace aare

include/aare/ClusterCollector.hpp

@@ -2,29 +2,31 @@
 #include <atomic>
 #include <thread>
 
-#include "aare/ProducerConsumerQueue.hpp"
-#include "aare/ClusterVector.hpp"
 #include "aare/ClusterFinderMT.hpp"
+#include "aare/ClusterVector.hpp"
+#include "aare/ProducerConsumerQueue.hpp"
 
 namespace aare {
 
-class ClusterCollector{
-        ProducerConsumerQueue<ClusterVector<int>>* m_source;
-        std::atomic<bool> m_stop_requested{false};
-        std::atomic<bool> m_stopped{true};
-        std::chrono::milliseconds m_default_wait{1};
-        std::thread m_thread;
-        std::vector<ClusterVector<int>> m_clusters;
+template <typename ClusterType,
+          typename = std::enable_if_t<is_cluster_v<ClusterType>>>
+class ClusterCollector {
+    ProducerConsumerQueue<ClusterVector<ClusterType>> *m_source;
+    std::atomic<bool> m_stop_requested{false};
+    std::atomic<bool> m_stopped{true};
+    std::chrono::milliseconds m_default_wait{1};
+    std::thread m_thread;
+    std::vector<ClusterVector<ClusterType>> m_clusters;
 
-    void process(){
+    void process() {
         m_stopped = false;
         fmt::print("ClusterCollector started\n");
-        while (!m_stop_requested || !m_source->isEmpty()) { 
-            if (ClusterVector<int> *clusters = m_source->frontPtr();
+        while (!m_stop_requested || !m_source->isEmpty()) {
+            if (ClusterVector<ClusterType> *clusters = m_source->frontPtr();
                 clusters != nullptr) {
                 m_clusters.push_back(std::move(*clusters));
                 m_source->popFront();
-            }else{
+            } else {
                 std::this_thread::sleep_for(m_default_wait);
             }
         }
@@ -32,21 +34,25 @@ class ClusterCollector{
         m_stopped = true;
     }
 
-    public:
-        ClusterCollector(ClusterFinderMT<uint16_t, double, int32_t>* source){
-            m_source = source->sink();
-            m_thread = std::thread(&ClusterCollector::process, this);
-        }
-        void stop(){
-            m_stop_requested = true;
-            m_thread.join();
-        }
-        std::vector<ClusterVector<int>> steal_clusters(){
-            if(!m_stopped){
-                throw std::runtime_error("ClusterCollector is still running");
-            }
-            return std::move(m_clusters);
+  public:
+    ClusterCollector(ClusterFinderMT<ClusterType, uint16_t, double> *source) {
+        m_source = source->sink();
+        m_thread =
+            std::thread(&ClusterCollector::process,
+                        this); // only one process does that so why isnt it
+                               // automatically written to m_cluster in collect
+                               // - instead of writing first to m_sink?
+    }
+    void stop() {
+        m_stop_requested = true;
+        m_thread.join();
+    }
+    std::vector<ClusterVector<ClusterType>> steal_clusters() {
+        if (!m_stopped) {
+            throw std::runtime_error("ClusterCollector is still running");
         }
+        return std::move(m_clusters);
+    }
 };
 
 } // namespace aare

include/aare/ClusterFile.hpp

@@ -1,51 +1,16 @@
 #pragma once
 
+#include "aare/Cluster.hpp"
 #include "aare/ClusterVector.hpp"
+#include "aare/GainMap.hpp"
 #include "aare/NDArray.hpp"
 #include "aare/defs.hpp"
 #include <filesystem>
 #include <fstream>
+#include <optional>
 
 namespace aare {
 
-struct Cluster3x3 {
-    int16_t x;
-    int16_t y;
-    int32_t data[9];
-};
-
-typedef enum {
-    cBottomLeft = 0,
-    cBottomRight = 1,
-    cTopLeft = 2,
-    cTopRight = 3
-} corner;
-
-typedef enum {
-    pBottomLeft = 0,
-    pBottom = 1,
-    pBottomRight = 2,
-    pLeft = 3,
-    pCenter = 4,
-    pRight = 5,
-    pTopLeft = 6,
-    pTop = 7,
-    pTopRight = 8
-} pixel;
-
-struct Eta2 {
-    double x;
-    double y;
-    corner c;
-};
-
-struct ClusterAnalysis {
-    uint32_t c;
-    int32_t tot;
-    double etax;
-    double etay;
-};
-
 /*
 Binary cluster file. Expects data to be layed out as:
 int32_t frame_number
@@ -56,6 +21,8 @@ uint32_t number_of_clusters
 ....
 */
 
+// TODO: change to support any type of clusters, e.g. header line with
+// clsuter_size_x, cluster_size_y,
 /**
  * @brief Class to read and write cluster files
  * Expects data to be laid out as:
@@ -63,16 +30,24 @@ uint32_t number_of_clusters
  *
  *       int32_t frame_number
  *       uint32_t number_of_clusters
- *       int16_t x, int16_t y, int32_t data[9] x number_of_clusters
+ *       int16_t x, int16_t y, int32_t data[9] * number_of_clusters
  *       int32_t frame_number
  *       uint32_t number_of_clusters
  *       etc.
  */
+template <typename ClusterType,
+          typename Enable = std::enable_if_t<is_cluster_v<ClusterType>>>
 class ClusterFile {
     FILE *fp{};
-    uint32_t m_num_left{};
-    size_t m_chunk_size{};
-    const std::string m_mode;
+    const std::string m_filename{};
+    uint32_t m_num_left{};    /*Number of photons left in frame*/
+    size_t m_chunk_size{};    /*Number of clusters to read at a time*/
+    std::string m_mode;       /*Mode to open the file in*/
+    std::optional<ROI> m_roi; /*Region of interest, will be applied if set*/
+    std::optional<NDArray<int32_t, 2>>
+        m_noise_map; /*Noise map to cut photons, will be applied if set*/
+    std::optional<InvertedGainMap> m_gain_map; /*Gain map to apply to the
+                                          clusters, will be applied if set*/
 
   public:
     /**
@@ -85,51 +60,390 @@ class ClusterFile {
      * @throws std::runtime_error if the file could not be opened
      */
     ClusterFile(const std::filesystem::path &fname, size_t chunk_size = 1000,
-                const std::string &mode = "r");
-    
-    
-    ~ClusterFile();
+                const std::string &mode = "r")
+
+        : m_filename(fname.string()), m_chunk_size(chunk_size), m_mode(mode) {
+
+        if (mode == "r") {
+            fp = fopen(m_filename.c_str(), "rb");
+            if (!fp) {
+                throw std::runtime_error("Could not open file for reading: " +
+                                         m_filename);
+            }
+        } else if (mode == "w") {
+            fp = fopen(m_filename.c_str(), "wb");
+            if (!fp) {
+                throw std::runtime_error("Could not open file for writing: " +
+                                         m_filename);
+            }
+        } else if (mode == "a") {
+            fp = fopen(m_filename.c_str(), "ab");
+            if (!fp) {
+                throw std::runtime_error("Could not open file for appending: " +
+                                         m_filename);
+            }
+        } else {
+            throw std::runtime_error("Unsupported mode: " + mode);
+        }
+    }
+
+    ~ClusterFile() { close(); }
 
     /**
-     * @brief Read n_clusters clusters from the file discarding frame numbers.
-     * If EOF is reached the returned vector will have less than n_clusters
-     * clusters
+     * @brief Read n_clusters clusters from the file discarding
+     * frame numbers. If EOF is reached the returned vector will
+     * have less than n_clusters clusters
      */
-    ClusterVector<int32_t> read_clusters(size_t n_clusters);
+    ClusterVector<ClusterType> read_clusters(size_t n_clusters) {
+        if (m_mode != "r") {
+            throw std::runtime_error("File not opened for reading");
+        }
+        if (m_noise_map || m_roi) {
+            return read_clusters_with_cut(n_clusters);
+        } else {
+            return read_clusters_without_cut(n_clusters);
+        }
+    }
 
     /**
-     * @brief Read a single frame from the file and return the clusters. The
-     * cluster vector will have the frame number set.
-     * @throws std::runtime_error if the file is not opened for reading or the file pointer not
-     * at the beginning of a frame
+     * @brief Read a single frame from the file and return the
+     * clusters. The cluster vector will have the frame number
+     * set.
+     * @throws std::runtime_error if the file is not opened for
+     * reading or the file pointer not at the beginning of a
+     * frame
      */
-    ClusterVector<int32_t> read_frame();
+    ClusterVector<ClusterType> read_frame() {
+        if (m_mode != "r") {
+            throw std::runtime_error(LOCATION + "File not opened for reading");
+        }
+        if (m_noise_map || m_roi) {
+            return read_frame_with_cut();
+        } else {
+            return read_frame_without_cut();
+        }
+    }
 
+    void write_frame(const ClusterVector<ClusterType> &clusters) {
+        if (m_mode != "w" && m_mode != "a") {
+            throw std::runtime_error("File not opened for writing");
+        }
 
-    void write_frame(const ClusterVector<int32_t> &clusters);
-    
-    // Need to be migrated to support NDArray and return a ClusterVector
-    // std::vector<Cluster3x3>
-    // read_cluster_with_cut(size_t n_clusters, double *noise_map, int nx, int ny);
+        int32_t frame_number = clusters.frame_number();
+        fwrite(&frame_number, sizeof(frame_number), 1, fp);
+        uint32_t n_clusters = clusters.size();
+        fwrite(&n_clusters, sizeof(n_clusters), 1, fp);
+        fwrite(clusters.data(), clusters.item_size(), clusters.size(), fp);
+    }
 
     /**
      * @brief Return the chunk size
      */
     size_t chunk_size() const { return m_chunk_size; }
-    
-    
+
     /**
-     * @brief Close the file. If not closed the file will be closed in the destructor
+     * @brief Set the region of interest to use when reading
+     * clusters. If set only clusters within the ROI will be
+     * read.
      */
-    void close();
+    void set_roi(ROI roi) { m_roi = roi; }
+
+    /**
+     * @brief Set the noise map to use when reading clusters. If
+     * set clusters below the noise level will be discarded.
+     * Selection criteria one of: Central pixel above noise,
+     * highest 2x2 sum above 2 * noise, total sum above 3 *
+     * noise.
+     */
+    void set_noise_map(const NDView<int32_t, 2> noise_map) {
+        m_noise_map = NDArray<int32_t, 2>(noise_map);
+    }
+
+    /**
+     * @brief Set the gain map to use when reading clusters. If set the gain map
+     * will be applied to the clusters that pass ROI and noise_map selection.
+     * The gain map is expected to be in ADU/energy.
+     */
+    void set_gain_map(const NDView<double, 2> gain_map) {
+        m_gain_map = InvertedGainMap(gain_map);
+    }
+
+    void set_gain_map(const InvertedGainMap &gain_map) {
+        m_gain_map = gain_map;
+    }
+
+    void set_gain_map(const InvertedGainMap &&gain_map) {
+        m_gain_map = gain_map;
+    }
+
+    /**
+     * @brief Close the file. If not closed the file will be
+     * closed in the destructor
+     */
+    void close() {
+        if (fp) {
+            fclose(fp);
+            fp = nullptr;
+        }
+    }
+
+    /** @brief Open the file in specific mode
+     *
+     */
+    void open(const std::string &mode) {
+        if (fp) {
+            close();
+        }
+
+        if (mode == "r") {
+            fp = fopen(m_filename.c_str(), "rb");
+            if (!fp) {
+                throw std::runtime_error("Could not open file for reading: " +
+                                         m_filename);
+            }
+            m_mode = "r";
+        } else if (mode == "w") {
+            fp = fopen(m_filename.c_str(), "wb");
+            if (!fp) {
+                throw std::runtime_error("Could not open file for writing: " +
+                                         m_filename);
+            }
+            m_mode = "w";
+        } else if (mode == "a") {
+            fp = fopen(m_filename.c_str(), "ab");
+            if (!fp) {
+                throw std::runtime_error("Could not open file for appending: " +
+                                         m_filename);
+            }
+            m_mode = "a";
+        } else {
+            throw std::runtime_error("Unsupported mode: " + mode);
+        }
+    }
+
+  private:
+    ClusterVector<ClusterType> read_clusters_with_cut(size_t n_clusters);
+    ClusterVector<ClusterType> read_clusters_without_cut(size_t n_clusters);
+    ClusterVector<ClusterType> read_frame_with_cut();
+    ClusterVector<ClusterType> read_frame_without_cut();
+    bool is_selected(ClusterType &cl);
+    ClusterType read_one_cluster();
 };
 
-int analyze_data(int32_t *data, int32_t *t2, int32_t *t3, char *quad,
-                 double *eta2x, double *eta2y, double *eta3x, double *eta3y);
-int analyze_cluster(Cluster3x3 &cl, int32_t *t2, int32_t *t3, char *quad,
-                    double *eta2x, double *eta2y, double *eta3x, double *eta3y);
+template <typename ClusterType, typename Enable>
+ClusterVector<ClusterType>
+ClusterFile<ClusterType, Enable>::read_clusters_without_cut(size_t n_clusters) {
+    if (m_mode != "r") {
+        throw std::runtime_error("File not opened for reading");
+    }
 
-NDArray<double, 2> calculate_eta2(ClusterVector<int> &clusters);
-Eta2 calculate_eta2(Cluster3x3 &cl);
+    ClusterVector<ClusterType> clusters(n_clusters);
+    clusters.resize(n_clusters);
+
+    int32_t iframe = 0; // frame number needs to be 4 bytes!
+    size_t nph_read = 0;
+    uint32_t nn = m_num_left;
+    uint32_t nph = m_num_left; // number of clusters in frame needs to be 4
+
+    auto buf = clusters.data();
+    // if there are photons left from previous frame read them first
+    if (nph) {
+        if (nph > n_clusters) {
+            // if we have more photons left in the frame then photons to
+            // read we read directly the requested number
+            nn = n_clusters;
+        } else {
+            nn = nph;
+        }
+        nph_read += fread((buf + nph_read), clusters.item_size(), nn, fp);
+        m_num_left = nph - nn; // write back the number of photons left
+    }
+
+    if (nph_read < n_clusters) {
+        // keep on reading frames and photons until reaching n_clusters
+        while (fread(&iframe, sizeof(iframe), 1, fp)) {
+            clusters.set_frame_number(iframe);
+            // read number of clusters in frame
+            if (fread(&nph, sizeof(nph), 1, fp)) {
+                if (nph > (n_clusters - nph_read))
+                    nn = n_clusters - nph_read;
+                else
+                    nn = nph;
+
+                nph_read +=
+                    fread((buf + nph_read), clusters.item_size(), nn, fp);
+                m_num_left = nph - nn;
+            }
+            if (nph_read >= n_clusters)
+                break;
+        }
+    }
+
+    // Resize the vector to the number o f clusters.
+    // No new allocation, only change bounds.
+    clusters.resize(nph_read);
+    if (m_gain_map)
+        m_gain_map->apply_gain_map(clusters);
+    return clusters;
+}
+
+template <typename ClusterType, typename Enable>
+ClusterVector<ClusterType>
+ClusterFile<ClusterType, Enable>::read_clusters_with_cut(size_t n_clusters) {
+    ClusterVector<ClusterType> clusters;
+    clusters.reserve(n_clusters);
+
+    // if there are photons left from previous frame read them first
+    if (m_num_left) {
+        while (m_num_left && clusters.size() < n_clusters) {
+            ClusterType c = read_one_cluster();
+            if (is_selected(c)) {
+                clusters.push_back(c);
+            }
+        }
+    }
+
+    // we did not have enough clusters left in the previous frame
+    // keep on reading frames until reaching n_clusters
+    if (clusters.size() < n_clusters) {
+        // sanity check
+        if (m_num_left) {
+            throw std::runtime_error(
+                LOCATION + "Entered second loop with clusters left\n");
+        }
+
+        int32_t frame_number = 0; // frame number needs to be 4 bytes!
+        while (fread(&frame_number, sizeof(frame_number), 1, fp)) {
+            if (fread(&m_num_left, sizeof(m_num_left), 1, fp)) {
+                clusters.set_frame_number(
+                    frame_number); // cluster vector will hold the last
+                                   // frame number
+                while (m_num_left && clusters.size() < n_clusters) {
+                    ClusterType c = read_one_cluster();
+                    if (is_selected(c)) {
+                        clusters.push_back(c);
+                    }
+                }
+            }
+
+            // we have enough clusters, break out of the outer while loop
+            if (clusters.size() >= n_clusters)
+                break;
+        }
+    }
+    if (m_gain_map)
+        m_gain_map->apply_gain_map(clusters);
+
+    return clusters;
+}
+
+template <typename ClusterType, typename Enable>
+ClusterType ClusterFile<ClusterType, Enable>::read_one_cluster() {
+    ClusterType c;
+    auto rc = fread(&c, sizeof(c), 1, fp);
+    if (rc != 1) {
+        throw std::runtime_error(LOCATION + "Could not read cluster");
+    }
+    --m_num_left;
+    return c;
+}
+
+template <typename ClusterType, typename Enable>
+ClusterVector<ClusterType>
+ClusterFile<ClusterType, Enable>::read_frame_without_cut() {
+    if (m_mode != "r") {
+        throw std::runtime_error("File not opened for reading");
+    }
+    if (m_num_left) {
+        throw std::runtime_error(
+            "There are still photons left in the last frame");
+    }
+    int32_t frame_number;
+    if (fread(&frame_number, sizeof(frame_number), 1, fp) != 1) {
+        throw std::runtime_error(LOCATION + "Could not read frame number");
+    }
+
+    int32_t n_clusters; // Saved as 32bit integer in the cluster file
+    if (fread(&n_clusters, sizeof(n_clusters), 1, fp) != 1) {
+        throw std::runtime_error(LOCATION +
+                                 "Could not read number of clusters");
+    }
+
+    ClusterVector<ClusterType> clusters(n_clusters);
+    clusters.set_frame_number(frame_number);
+
+    clusters.resize(n_clusters);
+
+    if (fread(clusters.data(), clusters.item_size(), n_clusters, fp) !=
+        static_cast<size_t>(n_clusters)) {
+        throw std::runtime_error(LOCATION + "Could not read clusters");
+    }
+
+    if (m_gain_map)
+        m_gain_map->apply_gain_map(clusters);
+    return clusters;
+}
+
+template <typename ClusterType, typename Enable>
+ClusterVector<ClusterType>
+ClusterFile<ClusterType, Enable>::read_frame_with_cut() {
+    if (m_mode != "r") {
+        throw std::runtime_error("File not opened for reading");
+    }
+    if (m_num_left) {
+        throw std::runtime_error(
+            "There are still photons left in the last frame");
+    }
+    int32_t frame_number;
+    if (fread(&frame_number, sizeof(frame_number), 1, fp) != 1) {
+        throw std::runtime_error("Could not read frame number");
+    }
+
+    if (fread(&m_num_left, sizeof(m_num_left), 1, fp) != 1) {
+        throw std::runtime_error("Could not read number of clusters");
+    }
+
+    ClusterVector<ClusterType> clusters;
+    clusters.reserve(m_num_left);
+    clusters.set_frame_number(frame_number);
+    while (m_num_left) {
+        ClusterType c = read_one_cluster();
+        if (is_selected(c)) {
+            clusters.push_back(c);
+        }
+    }
+    if (m_gain_map)
+        m_gain_map->apply_gain_map(clusters);
+    return clusters;
+}
+
+template <typename ClusterType, typename Enable>
+bool ClusterFile<ClusterType, Enable>::is_selected(ClusterType &cl) {
+    // Should fail fast
+    if (m_roi) {
+        if (!(m_roi->contains(cl.x, cl.y))) {
+            return false;
+        }
+    }
+
+    size_t cluster_center_index =
+        (ClusterType::cluster_size_x / 2) +
+        (ClusterType::cluster_size_y / 2) * ClusterType::cluster_size_x;
+
+    if (m_noise_map) {
+        auto sum_1x1 = cl.data[cluster_center_index]; // central pixel
+        auto sum_2x2 = cl.max_sum_2x2().first; // highest sum of 2x2 subclusters
+        auto total_sum = cl.sum();             // sum of all pixels
+
+        auto noise =
+            (*m_noise_map)(cl.y, cl.x); // TODO! check if this is correct
+        if (sum_1x1 <= noise || sum_2x2 <= 2 * noise ||
+            total_sum <= 3 * noise) {
+            return false;
+        }
+    }
+    // we passed all checks
+    return true;
+}
 
 } // namespace aare

include/aare/ClusterFileSink.hpp

@@ -3,35 +3,41 @@
 #include <filesystem>
 #include <thread>
 
-#include "aare/ProducerConsumerQueue.hpp"
-#include "aare/ClusterVector.hpp"
 #include "aare/ClusterFinderMT.hpp"
+#include "aare/ClusterVector.hpp"
+#include "aare/ProducerConsumerQueue.hpp"
 
-namespace aare{
+namespace aare {
 
-class ClusterFileSink{
-    ProducerConsumerQueue<ClusterVector<int>>* m_source;
+template <typename ClusterType,
+          typename = std::enable_if_t<is_cluster_v<ClusterType>>>
+class ClusterFileSink {
+    ProducerConsumerQueue<ClusterVector<ClusterType>> *m_source;
     std::atomic<bool> m_stop_requested{false};
     std::atomic<bool> m_stopped{true};
     std::chrono::milliseconds m_default_wait{1};
     std::thread m_thread;
     std::ofstream m_file;
 
-
-    void process(){
+    void process() {
         m_stopped = false;
         fmt::print("ClusterFileSink started\n");
-        while (!m_stop_requested || !m_source->isEmpty()) { 
-            if (ClusterVector<int> *clusters = m_source->frontPtr();
+        while (!m_stop_requested || !m_source->isEmpty()) {
+            if (ClusterVector<ClusterType> *clusters = m_source->frontPtr();
                 clusters != nullptr) {
                 // Write clusters to file
-                int32_t frame_number = clusters->frame_number(); //TODO! Should we store frame number already as int?
+                int32_t frame_number =
+                    clusters->frame_number(); // TODO! Should we store frame
+                                              // number already as int?
                 uint32_t num_clusters = clusters->size();
-                m_file.write(reinterpret_cast<const char*>(&frame_number), sizeof(frame_number));
-                m_file.write(reinterpret_cast<const char*>(&num_clusters), sizeof(num_clusters));
-                m_file.write(reinterpret_cast<const char*>(clusters->data()), clusters->size() * clusters->item_size());
+                m_file.write(reinterpret_cast<const char *>(&frame_number),
+                             sizeof(frame_number));
+                m_file.write(reinterpret_cast<const char *>(&num_clusters),
+                             sizeof(num_clusters));
+                m_file.write(reinterpret_cast<const char *>(clusters->data()),
+                             clusters->size() * clusters->item_size());
                 m_source->popFront();
-            }else{
+            } else {
                 std::this_thread::sleep_for(m_default_wait);
             }
         }
@@ -39,18 +45,18 @@ class ClusterFileSink{
         m_stopped = true;
     }
 
-    public:
-        ClusterFileSink(ClusterFinderMT<uint16_t, double, int32_t>* source, const std::filesystem::path& fname){
-            m_source = source->sink();
-            m_thread = std::thread(&ClusterFileSink::process, this);
-            m_file.open(fname, std::ios::binary);
-        }
-        void stop(){
-            m_stop_requested = true;
-            m_thread.join();
-            m_file.close();
-        }
+  public:
+    ClusterFileSink(ClusterFinderMT<ClusterType, uint16_t, double> *source,
+                    const std::filesystem::path &fname) {
+        m_source = source->sink();
+        m_thread = std::thread(&ClusterFileSink::process, this);
+        m_file.open(fname, std::ios::binary);
+    }
+    void stop() {
+        m_stop_requested = true;
+        m_thread.join();
+        m_file.close();
+    }
 };
 
-
 } // namespace aare

include/aare/ClusterFileV2.hpp

@@ -1,148 +0,0 @@
-#pragma once
-#include "aare/core/defs.hpp"
-#include <filesystem>
-#include <string>
-#include <fmt/format.h>
-
-namespace aare {
-struct ClusterHeader {
-    int32_t frame_number;
-    int32_t n_clusters;
-    std::string to_string() const {
-        return "frame_number: " + std::to_string(frame_number) + ", n_clusters: " + std::to_string(n_clusters);
-    }
-};
-
-struct ClusterV2_ {
-    int16_t x;
-    int16_t y;
-    std::array<int32_t, 9> data;
-    std::string to_string(bool detailed = false) const {
-        if (detailed) {
-            std::string data_str = "[";
-            for (auto &d : data) {
-                data_str += std::to_string(d) + ", ";
-            }
-            data_str += "]";
-            return "x: " + std::to_string(x) + ", y: " + std::to_string(y) + ", data: " + data_str;
-        }
-        return "x: " + std::to_string(x) + ", y: " + std::to_string(y);
-    }
-};
-
-struct ClusterV2 {
-    ClusterV2_ cluster;
-    int32_t frame_number;
-    std::string to_string() const {
-        return "frame_number: " + std::to_string(frame_number) + ", " + cluster.to_string();
-    }
-};
-
-/**
- * @brief
- * important not: fp always points to the clusters header and does not point to individual clusters
- *
- */
-class ClusterFileV2 {
-    std::filesystem::path m_fpath; 
-    std::string m_mode;
-    FILE *fp{nullptr};
-
-    void check_open(){
-        if (!fp)
-            throw std::runtime_error(fmt::format("File: {} not open", m_fpath.string()));
-    }
-
-  public:
-    ClusterFileV2(std::filesystem::path const &fpath, std::string const &mode): m_fpath(fpath), m_mode(mode) {
-        if (m_mode != "r" && m_mode != "w")
-            throw std::invalid_argument("mode must be 'r' or 'w'");
-        if (m_mode == "r" && !std::filesystem::exists(m_fpath))
-            throw std::invalid_argument("File does not exist");
-        if (mode == "r") {
-            fp = fopen(fpath.string().c_str(), "rb");
-        } else if (mode == "w") {
-            if (std::filesystem::exists(fpath)) {
-                fp = fopen(fpath.string().c_str(), "r+b");
-            } else {
-                fp = fopen(fpath.string().c_str(), "wb");
-            }
-        }
-        if (fp == nullptr) {
-            throw std::runtime_error("Failed to open file");
-        }
-    }
-    ~ClusterFileV2() { close(); }
-    std::vector<ClusterV2> read() {
-        check_open();
-
-        ClusterHeader header;
-        fread(&header, sizeof(ClusterHeader), 1, fp); 
-        std::vector<ClusterV2_> clusters_(header.n_clusters);
-        fread(clusters_.data(), sizeof(ClusterV2_), header.n_clusters, fp);
-        std::vector<ClusterV2> clusters;
-        for (auto &c : clusters_) {
-            ClusterV2 cluster;
-            cluster.cluster = std::move(c);
-            cluster.frame_number = header.frame_number;
-            clusters.push_back(cluster);
-        }
-
-        return clusters;
-    }
-    std::vector<std::vector<ClusterV2>> read(int n_frames) {
-        std::vector<std::vector<ClusterV2>> clusters;
-        for (int i = 0; i < n_frames; i++) {
-            clusters.push_back(read());
-        }
-        return clusters;
-    }
-
-    size_t write(std::vector<ClusterV2> const &clusters) {
-        check_open();
-        if (m_mode != "w")
-            throw std::runtime_error("File not opened in write mode");
-        if (clusters.empty())
-            return 0;
-
-        ClusterHeader header;
-        header.frame_number = clusters[0].frame_number;
-        header.n_clusters = clusters.size();
-        fwrite(&header, sizeof(ClusterHeader), 1, fp);
-        for (auto &c : clusters) {
-            fwrite(&c.cluster, sizeof(ClusterV2_), 1, fp);
-        }
-        return clusters.size();
-    }
-
-    size_t write(std::vector<std::vector<ClusterV2>> const &clusters) {
-        check_open();
-        if (m_mode != "w") 
-            throw std::runtime_error("File not opened in write mode");
-
-        size_t n_clusters = 0;
-        for (auto &c : clusters) {
-            n_clusters += write(c);
-        }
-        return n_clusters;
-    }
-
-    int seek_to_begin() { return fseek(fp, 0, SEEK_SET); }
-    int seek_to_end() { return fseek(fp, 0, SEEK_END); }
-
-    int32_t frame_number() {
-        auto pos = ftell(fp);
-        ClusterHeader header;
-        fread(&header, sizeof(ClusterHeader), 1, fp);
-        fseek(fp, pos, SEEK_SET);
-        return header.frame_number;
-    }
-
-    void close() {
-        if (fp) {
-            fclose(fp);
-            fp = nullptr;
-        }
-    }
-};
-} // namespace aare

include/aare/ClusterFinder.hpp

@@ -10,17 +10,19 @@
 
 namespace aare {
 
-template <typename FRAME_TYPE = uint16_t, typename PEDESTAL_TYPE = double,
-          typename CT = int32_t>
+template <typename ClusterType = Cluster<int32_t, 3, 3>,
+          typename FRAME_TYPE = uint16_t, typename PEDESTAL_TYPE = double>
 class ClusterFinder {
     Shape<2> m_image_size;
-    const int m_cluster_sizeX;
-    const int m_cluster_sizeY;
     const PEDESTAL_TYPE m_nSigma;
     const PEDESTAL_TYPE c2;
     const PEDESTAL_TYPE c3;
     Pedestal<PEDESTAL_TYPE> m_pedestal;
-    ClusterVector<CT> m_clusters;
+    ClusterVector<ClusterType> m_clusters;
+
+    static const uint8_t ClusterSizeX = ClusterType::cluster_size_x;
+    static const uint8_t ClusterSizeY = ClusterType::cluster_size_y;
+    using CT = typename ClusterType::value_type;
 
   public:
     /**
@@ -31,15 +33,12 @@ class ClusterFinder {
      * @param capacity initial capacity of the cluster vector
      *
      */
-    ClusterFinder(Shape<2> image_size, Shape<2> cluster_size,
-                  PEDESTAL_TYPE nSigma = 5.0, size_t capacity = 1000000)
-        : m_image_size(image_size), m_cluster_sizeX(cluster_size[0]),
-          m_cluster_sizeY(cluster_size[1]),
-          m_nSigma(nSigma),
-          c2(sqrt((m_cluster_sizeY + 1) / 2 * (m_cluster_sizeX + 1) / 2)),
-          c3(sqrt(m_cluster_sizeX * m_cluster_sizeY)),
-          m_pedestal(image_size[0], image_size[1]),
-          m_clusters(m_cluster_sizeX, m_cluster_sizeY, capacity) {};
+    ClusterFinder(Shape<2> image_size, PEDESTAL_TYPE nSigma = 5.0,
+                  size_t capacity = 1000000)
+        : m_image_size(image_size), m_nSigma(nSigma),
+          c2(sqrt((ClusterSizeY + 1) / 2 * (ClusterSizeX + 1) / 2)),
+          c3(sqrt(ClusterSizeX * ClusterSizeY)),
+          m_pedestal(image_size[0], image_size[1]), m_clusters(capacity) {};
 
     void push_pedestal_frame(NDView<FRAME_TYPE, 2> frame) {
         m_pedestal.push(frame);
@@ -56,23 +55,28 @@ class ClusterFinder {
      * same capacity as the old one
      *
      */
-    ClusterVector<CT> steal_clusters(bool realloc_same_capacity = false) {
-        ClusterVector<CT> tmp = std::move(m_clusters);
+    ClusterVector<ClusterType>
+    steal_clusters(bool realloc_same_capacity = false) {
+        ClusterVector<ClusterType> tmp = std::move(m_clusters);
         if (realloc_same_capacity)
-            m_clusters = ClusterVector<CT>(m_cluster_sizeX, m_cluster_sizeY,
-                                           tmp.capacity());
+            m_clusters = ClusterVector<ClusterType>(tmp.capacity());
         else
-            m_clusters = ClusterVector<CT>(m_cluster_sizeX, m_cluster_sizeY);
+            m_clusters = ClusterVector<ClusterType>{};
         return tmp;
     }
     void find_clusters(NDView<FRAME_TYPE, 2> frame, uint64_t frame_number = 0) {
         // // TODO! deal with even size clusters
         // // currently 3,3 -> +/- 1
         // //  4,4 -> +/- 2
-        int dy = m_cluster_sizeY / 2;
-        int dx = m_cluster_sizeX / 2;
+        int dy = ClusterSizeY / 2;
+        int dx = ClusterSizeX / 2;
+        int has_center_pixel_x =
+            ClusterSizeX %
+            2; // for even sized clusters there is no proper cluster center and
+               // even amount of pixels around the center
+        int has_center_pixel_y = ClusterSizeY % 2;
+
         m_clusters.set_frame_number(frame_number);
-        std::vector<CT> cluster_data(m_cluster_sizeX * m_cluster_sizeY);
         for (int iy = 0; iy < frame.shape(0); iy++) {
             for (int ix = 0; ix < frame.shape(1); ix++) {
 
@@ -87,8 +91,8 @@ class ClusterFinder {
                     continue; // NEGATIVE_PEDESTAL go to next pixel
                               // TODO! No pedestal update???
 
-                for (int ir = -dy; ir < dy + 1; ir++) {
-                    for (int ic = -dx; ic < dx + 1; ic++) {
+                for (int ir = -dy; ir < dy + has_center_pixel_y; ir++) {
+                    for (int ic = -dx; ic < dx + has_center_pixel_x; ic++) {
                         if (ix + ic >= 0 && ix + ic < frame.shape(1) &&
                             iy + ir >= 0 && iy + ir < frame.shape(0)) {
                             PEDESTAL_TYPE val =
@@ -109,27 +113,33 @@ class ClusterFinder {
                     // pass
                 } else {
                     // m_pedestal.push(iy, ix, frame(iy, ix));   // Safe option
-                    m_pedestal.push_fast(iy, ix, frame(iy, ix)); // Assume we have reached n_samples in the pedestal, slight performance improvement
-                    continue; // It was a pedestal value nothing to store
+                    m_pedestal.push_fast(
+                        iy, ix,
+                        frame(iy,
+                              ix)); // Assume we have reached n_samples in the
+                                    // pedestal, slight performance improvement
+                    continue;       // It was a pedestal value nothing to store
                 }
 
                 // Store cluster
                 if (value == max) {
-                    // Zero out the cluster data
-                    std::fill(cluster_data.begin(), cluster_data.end(), 0);
+                    ClusterType cluster{};
+                    cluster.x = ix;
+                    cluster.y = iy;
 
                     // Fill the cluster data since we have a photon to store
                     // It's worth redoing the look since most of the time we
                     // don't have a photon
                     int i = 0;
-                    for (int ir = -dy; ir < dy + 1; ir++) {
-                        for (int ic = -dx; ic < dx + 1; ic++) {
+                    for (int ir = -dy; ir < dy + has_center_pixel_y; ir++) {
+                        for (int ic = -dx; ic < dx + has_center_pixel_y; ic++) {
                             if (ix + ic >= 0 && ix + ic < frame.shape(1) &&
                                 iy + ir >= 0 && iy + ir < frame.shape(0)) {
                                 CT tmp =
                                     static_cast<CT>(frame(iy + ir, ix + ic)) -
-                                    m_pedestal.mean(iy + ir, ix + ic);
-                                cluster_data[i] =
+                                    static_cast<CT>(
+                                        m_pedestal.mean(iy + ir, ix + ic));
+                                cluster.data[i] =
                                     tmp; // Watch for out of bounds access
                                 i++;
                             }
@@ -137,9 +147,7 @@ class ClusterFinder {
                     }
 
                     // Add the cluster to the output ClusterVector
-                    m_clusters.push_back(
-                        ix, iy,
-                        reinterpret_cast<std::byte *>(cluster_data.data()));
+                    m_clusters.push_back(cluster);
                 }
             }
         }

include/aare/ClusterFinderMT.hpp

@@ -30,14 +30,17 @@ struct FrameWrapper {
  * @tparam PEDESTAL_TYPE type of the pedestal data
  * @tparam CT type of the cluster data
  */
-template <typename FRAME_TYPE = uint16_t, typename PEDESTAL_TYPE = double,
-          typename CT = int32_t>
+template <typename ClusterType = Cluster<int32_t, 3, 3>,
+          typename FRAME_TYPE = uint16_t, typename PEDESTAL_TYPE = double>
 class ClusterFinderMT {
+
+  protected:
+    using CT = typename ClusterType::value_type;
     size_t m_current_thread{0};
     size_t m_n_threads{0};
-    using Finder = ClusterFinder<FRAME_TYPE, PEDESTAL_TYPE, CT>;
+    using Finder = ClusterFinder<ClusterType, FRAME_TYPE, PEDESTAL_TYPE>;
     using InputQueue = ProducerConsumerQueue<FrameWrapper>;
-    using OutputQueue = ProducerConsumerQueue<ClusterVector<int>>;
+    using OutputQueue = ProducerConsumerQueue<ClusterVector<ClusterType>>;
     std::vector<std::unique_ptr<InputQueue>> m_input_queues;
     std::vector<std::unique_ptr<OutputQueue>> m_output_queues;
 
@@ -48,6 +51,7 @@ class ClusterFinderMT {
     std::thread m_collect_thread;
     std::chrono::milliseconds m_default_wait{1};
 
+  private:
     std::atomic<bool> m_stop_requested{false};
     std::atomic<bool> m_processing_threads_stopped{true};
 
@@ -66,7 +70,8 @@ class ClusterFinderMT {
                 switch (frame->type) {
                 case FrameType::DATA:
                     cf->find_clusters(frame->data.view(), frame->frame_number);
-                    m_output_queues[thread_id]->write(cf->steal_clusters(realloc_same_capacity));
+                    m_output_queues[thread_id]->write(
+                        cf->steal_clusters(realloc_same_capacity));
                     break;
 
                 case FrameType::PEDESTAL:
@@ -114,28 +119,32 @@ class ClusterFinderMT {
      * expected number of clusters in a frame per frame.
      * @param n_threads number of threads to use
      */
-    ClusterFinderMT(Shape<2> image_size, Shape<2> cluster_size,
-                    PEDESTAL_TYPE nSigma = 5.0, size_t capacity = 2000,
-                    size_t n_threads = 3)
+    ClusterFinderMT(Shape<2> image_size, PEDESTAL_TYPE nSigma = 5.0,
+                    size_t capacity = 2000, size_t n_threads = 3)
         : m_n_threads(n_threads) {
+
         for (size_t i = 0; i < n_threads; i++) {
             m_cluster_finders.push_back(
-                std::make_unique<ClusterFinder<FRAME_TYPE, PEDESTAL_TYPE, CT>>(
-                    image_size, cluster_size, nSigma, capacity));
+                std::make_unique<
+                    ClusterFinder<ClusterType, FRAME_TYPE, PEDESTAL_TYPE>>(
+                    image_size, nSigma, capacity));
         }
         for (size_t i = 0; i < n_threads; i++) {
             m_input_queues.emplace_back(std::make_unique<InputQueue>(200));
             m_output_queues.emplace_back(std::make_unique<OutputQueue>(200));
         }
-        //TODO! Should we start automatically?
+        // TODO! Should we start automatically?
         start();
     }
 
     /**
      * @brief Return the sink queue where all the clusters are collected
-     * @warning You need to empty this queue otherwise the cluster finder will wait forever
+     * @warning You need to empty this queue otherwise the cluster finder will
+     * wait forever
      */
-    ProducerConsumerQueue<ClusterVector<int>> *sink() { return &m_sink; }
+    ProducerConsumerQueue<ClusterVector<ClusterType>> *sink() {
+        return &m_sink;
+    }
 
     /**
      * @brief Start all processing threads

include/aare/ClusterVector.hpp

@@ -1,4 +1,5 @@
 #pragma once
+#include "aare/Cluster.hpp" //TODO maybe store in seperate file !!!
 #include <algorithm>
 #include <array>
 #include <cstddef>
@@ -8,268 +9,162 @@
 
 #include <fmt/core.h>
 
+#include "aare/Cluster.hpp"
+#include "aare/NDView.hpp"
+
 namespace aare {
 
+template <typename ClusterType,
+          typename = std::enable_if_t<is_cluster_v<ClusterType>>>
+class ClusterVector; // Forward declaration
+
 /**
- * @brief ClusterVector is a container for clusters of various sizes. It uses a
- * contiguous memory buffer to store the clusters. It is templated on the data
- * type and the coordinate type of the clusters.
+ * @brief ClusterVector is a container for clusters of various sizes. It
+ * uses a contiguous memory buffer to store the clusters. It is templated on
+ * the data type and the coordinate type of the clusters.
  * @note push_back can invalidate pointers to elements in the container
- * @warning ClusterVector is currently move only to catch unintended copies, but
- * this might change since there are probably use cases where copying is needed.
+ * @warning ClusterVector is currently move only to catch unintended copies,
+ * but this might change since there are probably use cases where copying is
+ * needed.
  * @tparam T data type of the pixels in the cluster
  * @tparam CoordType data type of the x and y coordinates of the cluster
  * (normally int16_t)
  */
-template <typename T, typename CoordType = int16_t> class ClusterVector {
-    using value_type = T;
-    size_t m_cluster_size_x;
-    size_t m_cluster_size_y;
-    std::byte *m_data{};
-    size_t m_size{0};
-    size_t m_capacity;
-    uint64_t m_frame_number{0}; // TODO! Check frame number size and type
-    /*
-    Format string used in the python bindings to create a numpy
-    array from the buffer
-    = - native byte order
-    h - short
-    d - double
-    i - int
-    */
-    constexpr static char m_fmt_base[] = "=h:x:\nh:y:\n({},{}){}:data:";
+template <typename T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType>
+class ClusterVector<Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>> {
+
+    std::vector<Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>> m_data{};
+    int32_t m_frame_number{0}; // TODO! Check frame number size and type
 
   public:
+    using value_type = T;
+    using ClusterType = Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>;
+
     /**
      * @brief Construct a new ClusterVector object
-     * @param cluster_size_x size of the cluster in x direction
-     * @param cluster_size_y size of the cluster in y direction
      * @param capacity initial capacity of the buffer in number of clusters
      * @param frame_number frame number of the clusters. Default is 0, which is
      * also used to indicate that the clusters come from many frames
      */
-    ClusterVector(size_t cluster_size_x = 3, size_t cluster_size_y = 3,
-                  size_t capacity = 1024, uint64_t frame_number = 0)
-        : m_cluster_size_x(cluster_size_x), m_cluster_size_y(cluster_size_y),
-          m_capacity(capacity), m_frame_number(frame_number) {
-        allocate_buffer(capacity);
+    ClusterVector(size_t capacity = 1024, uint64_t frame_number = 0)
+        : m_frame_number(frame_number) {
+        m_data.reserve(capacity);
     }
 
-    ~ClusterVector() { delete[] m_data; }
-
     // Move constructor
     ClusterVector(ClusterVector &&other) noexcept
-        : m_cluster_size_x(other.m_cluster_size_x),
-          m_cluster_size_y(other.m_cluster_size_y), m_data(other.m_data),
-          m_size(other.m_size), m_capacity(other.m_capacity),
-          m_frame_number(other.m_frame_number) {
-        other.m_data = nullptr;
-        other.m_size = 0;
-        other.m_capacity = 0;
+        : m_data(other.m_data), m_frame_number(other.m_frame_number) {
+        other.m_data.clear();
     }
 
     // Move assignment operator
     ClusterVector &operator=(ClusterVector &&other) noexcept {
         if (this != &other) {
-            delete[] m_data;
-            m_cluster_size_x = other.m_cluster_size_x;
-            m_cluster_size_y = other.m_cluster_size_y;
             m_data = other.m_data;
-            m_size = other.m_size;
-            m_capacity = other.m_capacity;
             m_frame_number = other.m_frame_number;
-            other.m_data = nullptr;
-            other.m_size = 0;
-            other.m_capacity = 0;
+            other.m_data.clear();
             other.m_frame_number = 0;
         }
         return *this;
     }
 
-    /**
-     * @brief Reserve space for at least capacity clusters
-     * @param capacity number of clusters to reserve space for
-     * @note If capacity is less than the current capacity, the function does
-     * nothing.
-     */
-    void reserve(size_t capacity) {
-        if (capacity > m_capacity) {
-            allocate_buffer(capacity);
-        }
-    }
-
-    /**
-     * @brief Add a cluster to the vector
-     * @param x x-coordinate of the cluster
-     * @param y y-coordinate of the cluster
-     * @param data pointer to the data of the cluster
-     * @warning The data pointer must point to a buffer of size cluster_size_x *
-     * cluster_size_y * sizeof(T)
-     */
-    void push_back(CoordType x, CoordType y, const std::byte *data) {
-        if (m_size == m_capacity) {
-            allocate_buffer(m_capacity * 2);
-        }
-        std::byte *ptr = element_ptr(m_size);
-        *reinterpret_cast<CoordType *>(ptr) = x;
-        ptr += sizeof(CoordType);
-        *reinterpret_cast<CoordType *>(ptr) = y;
-        ptr += sizeof(CoordType);
-
-        std::copy(data, data + m_cluster_size_x * m_cluster_size_y * sizeof(T),
-                  ptr);
-        m_size++;
-    }
-    ClusterVector &operator+=(const ClusterVector &other) {
-        if (m_size + other.m_size > m_capacity) {
-            allocate_buffer(m_capacity + other.m_size);
-        }
-        std::copy(other.m_data, other.m_data + other.m_size * item_size(),
-                  m_data + m_size * item_size());
-        m_size += other.m_size;
-        return *this;
-    }
-
     /**
      * @brief Sum the pixels in each cluster
      * @return std::vector<T> vector of sums for each cluster
      */
     std::vector<T> sum() {
-        std::vector<T> sums(m_size);
-        const size_t stride = item_size();
-        const size_t n_pixels = m_cluster_size_x * m_cluster_size_y;
-        std::byte *ptr = m_data + 2 * sizeof(CoordType); // skip x and y
+        std::vector<T> sums(m_data.size());
+
+        std::transform(
+            m_data.begin(), m_data.end(), sums.begin(),
+            [](const ClusterType &cluster) { return cluster.sum(); });
 
-        for (size_t i = 0; i < m_size; i++) {
-            sums[i] =
-                std::accumulate(reinterpret_cast<T *>(ptr),
-                                reinterpret_cast<T *>(ptr) + n_pixels, T{});
-            ptr += stride;
-        }
         return sums;
     }
 
     /**
-     * @brief Return the maximum sum of the 2x2 subclusters in each cluster
+     * @brief Sum the pixels in the 2x2 subcluster with the biggest pixel sum in
+     * each cluster
      * @return std::vector<T> vector of sums for each cluster
-     * @throws std::runtime_error if the cluster size is not 3x3
-     * @warning Only 3x3 clusters are supported for the 2x2 sum.
      */
     std::vector<T> sum_2x2() {
-        std::vector<T> sums(m_size);
-        const size_t stride = item_size();
+        std::vector<T> sums_2x2(m_data.size());
 
-        if (m_cluster_size_x != 3 || m_cluster_size_y != 3) {
-            throw std::runtime_error(
-                "Only 3x3 clusters are supported for the 2x2 sum.");
-        }
-        std::byte *ptr = m_data + 2 * sizeof(CoordType); // skip x and y
+        std::transform(m_data.begin(), m_data.end(), sums_2x2.begin(),
+                       [](const ClusterType &cluster) {
+                           return cluster.max_sum_2x2().first;
+                       });
 
-        for (size_t i = 0; i < m_size; i++) {
-            std::array<T, 4> total;
-            auto T_ptr = reinterpret_cast<T *>(ptr);
-            total[0] = T_ptr[0] + T_ptr[1] + T_ptr[3] + T_ptr[4];
-            total[1] = T_ptr[1] + T_ptr[2] + T_ptr[4] + T_ptr[5];
-            total[2] = T_ptr[3] + T_ptr[4] + T_ptr[6] + T_ptr[7];
-            total[3] = T_ptr[4] + T_ptr[5] + T_ptr[7] + T_ptr[8];
+        return sums_2x2;
+    }
 
-            sums[i] = *std::max_element(total.begin(), total.end());
-            ptr += stride;
-        }
+    /**
+     * @brief Reserve space for at least capacity clusters
+     * @param capacity number of clusters to reserve space for
+     * @note If capacity is less than the current capacity, the function does
+     * nothing.
+     */
+    void reserve(size_t capacity) { m_data.reserve(capacity); }
 
-        return sums;
+    void resize(size_t size) { m_data.resize(size); }
+
+    void push_back(const ClusterType &cluster) { m_data.push_back(cluster); }
+
+    ClusterVector &operator+=(const ClusterVector &other) {
+        m_data.insert(m_data.end(), other.begin(), other.end());
+
+        return *this;
     }
 
     /**
      * @brief Return the number of clusters in the vector
      */
-    size_t size() const { return m_size; }
+    size_t size() const { return m_data.size(); }
+
+    uint8_t cluster_size_x() const { return ClusterSizeX; }
+
+    uint8_t cluster_size_y() const { return ClusterSizeY; }
 
     /**
      * @brief Return the capacity of the buffer in number of clusters. This is
      * the number of clusters that can be stored in the current buffer without
      * reallocation.
      */
-    size_t capacity() const { return m_capacity; }
+    size_t capacity() const { return m_data.capacity(); }
+
+    const auto begin() const { return m_data.begin(); }
+
+    const auto end() const { return m_data.end(); }
 
     /**
      * @brief Return the size in bytes of a single cluster
      */
     size_t item_size() const {
-        return 2 * sizeof(CoordType) +
-               m_cluster_size_x * m_cluster_size_y * sizeof(T);
+        return sizeof(ClusterType); // 2 * sizeof(CoordType) + ClusterSizeX *
+                                    // ClusterSizeY * sizeof(T);
     }
 
-    /**
-     * @brief Return the offset in bytes for the i-th cluster
-     */
-    size_t element_offset(size_t i) const { return item_size() * i; }
-
-    /**
-     * @brief Return a pointer to the i-th cluster
-     */
-    std::byte *element_ptr(size_t i) { return m_data + element_offset(i); }
-
-    /**
-     * @brief Return a pointer to the i-th cluster
-     */
-    const std::byte *element_ptr(size_t i) const {
-        return m_data + element_offset(i);
-    }
-
-    size_t cluster_size_x() const { return m_cluster_size_x; }
-    size_t cluster_size_y() const { return m_cluster_size_y; }
-
-    std::byte *data() { return m_data; }
-    std::byte const *data() const { return m_data; }
+    ClusterType *data() { return m_data.data(); }
+    ClusterType const *data() const { return m_data.data(); }
 
     /**
      * @brief Return a reference to the i-th cluster casted to type V
      * @tparam V type of the cluster
      */
-    template <typename V> V &at(size_t i) {
-        return *reinterpret_cast<V *>(element_ptr(i));
-    }
+    ClusterType &operator[](size_t i) { return m_data[i]; }
 
-    const std::string_view fmt_base() const {
-        // TODO! how do we match on coord_t?
-        return m_fmt_base;
-    }
+    const ClusterType &operator[](size_t i) const { return m_data[i]; }
 
     /**
      * @brief Return the frame number of the clusters. 0 is used to indicate
      * that the clusters come from many frames
      */
-    uint64_t frame_number() const { return m_frame_number; }
+    int32_t frame_number() const { return m_frame_number; }
 
-    void set_frame_number(uint64_t frame_number) {
+    void set_frame_number(int32_t frame_number) {
         m_frame_number = frame_number;
     }
-
-    /**
-     * @brief Resize the vector to contain new_size clusters. If new_size is
-     * greater than the current capacity, a new buffer is allocated. If the size
-     * is smaller no memory is freed, size is just updated.
-     * @param new_size new size of the vector
-     * @warning The additional clusters are not initialized
-     */
-    void resize(size_t new_size) {
-        // TODO! Should we initialize the new clusters?
-        if (new_size > m_capacity) {
-            allocate_buffer(new_size);
-        }
-        m_size = new_size;
-    }
-
-  private:
-    void allocate_buffer(size_t new_capacity) {
-        size_t num_bytes = item_size() * new_capacity;
-        std::byte *new_data = new std::byte[num_bytes]{};
-        std::copy(m_data, m_data + item_size() * m_size, new_data);
-        delete[] m_data;
-        m_data = new_data;
-        m_capacity = new_capacity;
-    }
 };
 
 } // namespace aare

include/aare/FilePtr.hpp

@@ -0,0 +1,30 @@
+#pragma once
+#include <cstdio>
+#include <filesystem>
+
+namespace aare {
+
+/**
+ * \brief RAII wrapper for FILE pointer
+ */
+class FilePtr {
+    FILE *fp_{nullptr};
+
+  public:
+    FilePtr() = default;
+    FilePtr(const std::filesystem::path& fname, const std::string& mode);
+    FilePtr(const FilePtr &) = delete;            // we don't want a copy
+    FilePtr &operator=(const FilePtr &) = delete; // since we handle a resource
+    FilePtr(FilePtr &&other);
+    FilePtr &operator=(FilePtr &&other);
+    FILE *get();
+    int64_t tell();
+    void seek(int64_t offset, int whence = SEEK_SET) {
+        if (fseek(fp_, offset, whence) != 0)
+            throw std::runtime_error("Error seeking in file");
+    }
+    std::string error_msg();
+    ~FilePtr();
+};
+
+} // namespace aare

include/aare/Fit.hpp

@@ -17,6 +17,14 @@ NDArray<double, 1> pol1(NDView<double, 1> x, NDView<double, 1> par);
 
 } // namespace func
 
+
+/**
+ * @brief Estimate the initial parameters for a Gaussian fit
+ */
+std::array<double, 3> gaus_init_par(const NDView<double, 1> x, const NDView<double, 1> y);
+
+std::array<double, 2> pol1_init_par(const NDView<double, 1> x, const NDView<double, 1> y);
+
 static constexpr int DEFAULT_NUM_THREADS = 4;
 
 /**
@@ -33,7 +41,11 @@ NDArray<double, 1> fit_gaus(NDView<double, 1> x, NDView<double, 1> y);
  * @param y y vales, layout [row, col, values]
  * @param n_threads number of threads to use
  */
-NDArray<double, 3> fit_gaus(NDView<double, 1> x, NDView<double, 3> y, int n_threads = DEFAULT_NUM_THREADS);
+
+NDArray<double, 3> fit_gaus(NDView<double, 1> x, NDView<double, 3> y,
+                            int n_threads = DEFAULT_NUM_THREADS);
+
+
 
 
 /**
@@ -45,10 +57,12 @@ NDArray<double, 3> fit_gaus(NDView<double, 1> x, NDView<double, 3> y, int n_thre
  * @param par_err_out output error parameters
  */
 void fit_gaus(NDView<double, 1> x, NDView<double, 1> y, NDView<double, 1> y_err,
-              NDView<double, 1> par_out, NDView<double, 1> par_err_out);
+              NDView<double, 1> par_out, NDView<double, 1> par_err_out,
+              double& chi2);
 
 /**
- * @brief Fit a 1D Gaussian to each pixel with error estimates. Data layout [row, col, values]
+ * @brief Fit a 1D Gaussian to each pixel with error estimates. Data layout
+ * [row, col, values]
  * @param x x values
  * @param y y vales, layout [row, col, values]
  * @param y_err error in y, layout [row, col, values]
@@ -57,20 +71,22 @@ void fit_gaus(NDView<double, 1> x, NDView<double, 1> y, NDView<double, 1> y_err,
  * @param n_threads number of threads to use
  */
 void fit_gaus(NDView<double, 1> x, NDView<double, 3> y, NDView<double, 3> y_err,
-              NDView<double, 3> par_out, NDView<double, 3> par_err_out, int n_threads = DEFAULT_NUM_THREADS);
-
+              NDView<double, 3> par_out, NDView<double, 3> par_err_out, NDView<double, 2> chi2_out,
+              int n_threads = DEFAULT_NUM_THREADS
+              );
 
 NDArray<double, 1> fit_pol1(NDView<double, 1> x, NDView<double, 1> y);
 
-NDArray<double, 3> fit_pol1(NDView<double, 1> x, NDView<double, 3> y, int n_threads = DEFAULT_NUM_THREADS);
+NDArray<double, 3> fit_pol1(NDView<double, 1> x, NDView<double, 3> y,
+                            int n_threads = DEFAULT_NUM_THREADS);
 
-void fit_pol1(NDView<double, 1> x, NDView<double, 1> y,
-                NDView<double, 1> y_err, NDView<double, 1> par_out,
-                NDView<double, 1> par_err_out);
+void fit_pol1(NDView<double, 1> x, NDView<double, 1> y, NDView<double, 1> y_err,
+              NDView<double, 1> par_out, NDView<double, 1> par_err_out, double& chi2);
+
+// TODO! not sure we need to offer the different version in C++
+void fit_pol1(NDView<double, 1> x, NDView<double, 3> y, NDView<double, 3> y_err,
+              NDView<double, 3> par_out, NDView<double, 3> par_err_out,NDView<double, 2> chi2_out,
+              int n_threads = DEFAULT_NUM_THREADS);
 
-//TODO! not sure we need to offer the different version in C++
-void fit_pol1(NDView<double, 1> x, NDView<double, 3> y,
-                NDView<double, 3> y_err, NDView<double, 3> par_out,
-                NDView<double, 3> par_err_out, int n_threads = DEFAULT_NUM_THREADS);
 
 } // namespace aare

include/aare/GainMap.hpp

@@ -0,0 +1,68 @@
+/************************************************
+ * @file GainMap.hpp
+ * @short function to apply gain map of image size to a vector of clusters -
+ *note stored gainmap is inverted for efficient aaplication to images
+ ***********************************************/
+
+#pragma once
+#include "aare/Cluster.hpp"
+#include "aare/ClusterVector.hpp"
+#include "aare/NDArray.hpp"
+#include "aare/NDView.hpp"
+#include <memory>
+
+namespace aare {
+
+class InvertedGainMap {
+
+  public:
+    explicit InvertedGainMap(const NDArray<double, 2> &gain_map)
+        : m_gain_map(gain_map) {
+        for (auto &item : m_gain_map) {
+            item = 1.0 / item;
+        }
+    };
+
+    explicit InvertedGainMap(const NDView<double, 2> gain_map) {
+        m_gain_map = NDArray<double, 2>(gain_map);
+        for (auto &item : m_gain_map) {
+            item = 1.0 / item;
+        }
+    }
+
+    template <typename ClusterType,
+              typename = std::enable_if_t<is_cluster_v<ClusterType>>>
+    void apply_gain_map(ClusterVector<ClusterType> &clustervec) {
+        // in principle we need to know the size of the image for this lookup
+        size_t ClusterSizeX = clustervec.cluster_size_x();
+        size_t ClusterSizeY = clustervec.cluster_size_y();
+
+        using T = typename ClusterVector<ClusterType>::value_type;
+
+        int64_t index_cluster_center_x = ClusterSizeX / 2;
+        int64_t index_cluster_center_y = ClusterSizeY / 2;
+        for (size_t i = 0; i < clustervec.size(); i++) {
+            auto &cl = clustervec[i];
+
+            if (cl.x > 0 && cl.y > 0 && cl.x < m_gain_map.shape(1) - 1 &&
+                cl.y < m_gain_map.shape(0) - 1) {
+                for (size_t j = 0; j < ClusterSizeX * ClusterSizeY; j++) {
+                    size_t x = cl.x + j % ClusterSizeX - index_cluster_center_x;
+                    size_t y = cl.y + j / ClusterSizeX - index_cluster_center_y;
+                    cl.data[j] = static_cast<T>(
+                        static_cast<double>(cl.data[j]) *
+                        m_gain_map(
+                            y, x)); // cast after conversion to keep precision
+                }
+            } else {
+                // clear edge clusters
+                cl.data.fill(0);
+            }
+        }
+    }
+
+  private:
+    NDArray<double, 2> m_gain_map{};
+};
+
+} // end of namespace aare

include/aare/Interpolator.hpp

@@ -0,0 +1,130 @@
+#pragma once
+
+#include "aare/CalculateEta.hpp"
+#include "aare/Cluster.hpp"
+#include "aare/ClusterFile.hpp" //Cluster_3x3
+#include "aare/ClusterVector.hpp"
+#include "aare/NDArray.hpp"
+#include "aare/NDView.hpp"
+#include "aare/algorithm.hpp"
+
+namespace aare {
+
+struct Photon {
+    double x;
+    double y;
+    double energy;
+};
+
+class Interpolator {
+    NDArray<double, 3> m_ietax;
+    NDArray<double, 3> m_ietay;
+
+    NDArray<double, 1> m_etabinsx;
+    NDArray<double, 1> m_etabinsy;
+    NDArray<double, 1> m_energy_bins;
+
+  public:
+    Interpolator(NDView<double, 3> etacube, NDView<double, 1> xbins,
+                 NDView<double, 1> ybins, NDView<double, 1> ebins);
+    NDArray<double, 3> get_ietax() { return m_ietax; }
+    NDArray<double, 3> get_ietay() { return m_ietay; }
+
+    template <typename ClusterType,
+              typename Eanble = std::enable_if_t<is_cluster_v<ClusterType>>>
+    std::vector<Photon> interpolate(const ClusterVector<ClusterType> &clusters);
+};
+
+// TODO: generalize to support any clustertype!!! otherwise add std::enable_if_t
+// to only take Cluster2x2 and Cluster3x3
+template <typename ClusterType, typename Enable>
+std::vector<Photon>
+Interpolator::interpolate(const ClusterVector<ClusterType> &clusters) {
+    std::vector<Photon> photons;
+    photons.reserve(clusters.size());
+
+    if (clusters.cluster_size_x() == 3 || clusters.cluster_size_y() == 3) {
+        for (const ClusterType &cluster : clusters) {
+
+            auto eta = calculate_eta2(cluster);
+
+            Photon photon;
+            photon.x = cluster.x;
+            photon.y = cluster.y;
+            photon.energy = eta.sum;
+
+            // auto ie = nearest_index(m_energy_bins, photon.energy)-1;
+            // auto ix = nearest_index(m_etabinsx, eta.x)-1;
+            // auto iy = nearest_index(m_etabinsy, eta.y)-1;
+            // Finding the index of the last element that is smaller
+            // should work fine as long as we have many bins
+            auto ie = last_smaller(m_energy_bins, photon.energy);
+            auto ix = last_smaller(m_etabinsx, eta.x);
+            auto iy = last_smaller(m_etabinsy, eta.y);
+
+            // fmt::print("ex: {}, ix: {}, iy: {}\n", ie, ix, iy);
+
+            double dX, dY;
+            // cBottomLeft = 0,
+            // cBottomRight = 1,
+            // cTopLeft = 2,
+            // cTopRight = 3
+            switch (static_cast<corner>(eta.c)) {
+            case corner::cTopLeft:
+                dX = -1.;
+                dY = 0;
+                break;
+            case corner::cTopRight:;
+                dX = 0;
+                dY = 0;
+                break;
+            case corner::cBottomLeft:
+                dX = -1.;
+                dY = -1.;
+                break;
+            case corner::cBottomRight:
+                dX = 0.;
+                dY = -1.;
+                break;
+            }
+            photon.x += m_ietax(ix, iy, ie) * 2 + dX;
+            photon.y += m_ietay(ix, iy, ie) * 2 + dY;
+            photons.push_back(photon);
+        }
+    } else if (clusters.cluster_size_x() == 2 ||
+               clusters.cluster_size_y() == 2) {
+        for (const ClusterType &cluster : clusters) {
+            auto eta = calculate_eta2(cluster);
+
+            Photon photon;
+            photon.x = cluster.x;
+            photon.y = cluster.y;
+            photon.energy = eta.sum;
+
+            // Now do some actual interpolation.
+            // Find which energy bin the cluster is in
+            //  auto ie = nearest_index(m_energy_bins, photon.energy)-1;
+            //  auto ix = nearest_index(m_etabinsx, eta.x)-1;
+            //  auto iy = nearest_index(m_etabinsy, eta.y)-1;
+            // Finding the index of the last element that is smaller
+            // should work fine as long as we have many bins
+            auto ie = last_smaller(m_energy_bins, photon.energy);
+            auto ix = last_smaller(m_etabinsx, eta.x);
+            auto iy = last_smaller(m_etabinsy, eta.y);
+
+            photon.x += m_ietax(ix, iy, ie) *
+                        2; // eta goes between 0 and 1 but we could move the hit
+                           // anywhere in the 2x2
+            photon.y += m_ietay(ix, iy, ie) * 2;
+            photons.push_back(photon);
+        }
+
+    } else {
+        throw std::runtime_error(
+            "Only 3x3 and 2x2 clusters are supported for interpolation");
+    }
+
+    return photons;
+}
+
+} // namespace aare

include/aare/JungfrauDataFile.hpp

@@ -0,0 +1,106 @@
+#pragma once
+#include <cstdint>
+#include <filesystem>
+#include <vector>
+
+#include "aare/FilePtr.hpp"
+#include "aare/defs.hpp"
+#include "aare/NDArray.hpp"
+#include "aare/FileInterface.hpp"
+namespace aare {
+
+
+struct JungfrauDataHeader{
+    uint64_t framenum;
+    uint64_t bunchid;
+};
+
+class JungfrauDataFile : public FileInterface {
+
+    size_t m_rows{};     //!< number of rows in the image, from find_frame_size();
+    size_t m_cols{};     //!< number of columns in the image, from find_frame_size();
+    size_t m_bytes_per_frame{}; //!< number of bytes per frame excluding header
+    size_t m_total_frames{};    //!< total number of frames in the series of files
+    size_t m_offset{}; //!< file index of the first file, allow starting at non zero file
+    size_t m_current_file_index{};   //!< The index of the open file
+    size_t m_current_frame_index{};  //!< The index of the current frame (with reference to all files)
+
+    std::vector<size_t> m_last_frame_in_file{}; //!< Used for seeking to the correct file
+    std::filesystem::path m_path; //!< path to the files
+    std::string m_base_name;      //!< base name used for formatting file names
+
+    FilePtr m_fp; //!< RAII wrapper for a FILE* 
+
+    
+    using pixel_type = uint16_t;
+    static constexpr size_t header_size = sizeof(JungfrauDataHeader); 
+    static constexpr size_t n_digits_in_file_index = 6; //!< to format file names
+
+  public:
+    JungfrauDataFile(const std::filesystem::path &fname);
+
+    std::string base_name() const;      //!< get the base name of the file (without path and extension)
+    size_t bytes_per_frame() override;           
+    size_t pixels_per_frame() override;    
+    size_t bytes_per_pixel() const;  
+    size_t bitdepth() const override;
+    void seek(size_t frame_index) override;  //!< seek to the given frame index (note not byte offset)
+    size_t tell() override;            //!< get the frame index of the file pointer
+    size_t total_frames() const override;
+    size_t rows() const override;
+    size_t cols() const override;
+    std::array<ssize_t,2> shape() const;
+    size_t n_files() const; //!< get the number of files in the series. 
+
+    // Extra functions needed for FileInterface
+    Frame read_frame() override;
+    Frame read_frame(size_t frame_number) override;
+    std::vector<Frame> read_n(size_t n_frames=0) override;
+    void read_into(std::byte *image_buf) override;
+    void read_into(std::byte *image_buf, size_t n_frames) override;
+    size_t frame_number(size_t frame_index) override;
+    DetectorType detector_type() const override;
+
+    /**
+     * @brief Read a single frame from the file into the given buffer.  
+     * @param image_buf buffer to read the frame into. (Note the caller is responsible for allocating the buffer)
+     * @param header pointer to a JungfrauDataHeader or nullptr to skip header)
+     */
+    void read_into(std::byte *image_buf, JungfrauDataHeader *header = nullptr);
+
+    /**
+     * @brief Read a multiple frames from the file into the given buffer.  
+     * @param image_buf buffer to read the frame into. (Note the caller is responsible for allocating the buffer)
+     * @param n_frames number of frames to read
+     * @param header pointer to a JungfrauDataHeader or nullptr to skip header)
+     */
+    void read_into(std::byte *image_buf, size_t n_frames, JungfrauDataHeader *header = nullptr);
+    
+    /** 
+     * @brief Read a single frame from the file into the given NDArray
+     * @param image NDArray to read the frame into.
+     */
+    void read_into(NDArray<uint16_t>* image, JungfrauDataHeader* header = nullptr);
+
+    JungfrauDataHeader read_header();
+    std::filesystem::path current_file() const { return fpath(m_current_file_index+m_offset); }
+
+    
+    private:
+    /**
+     * @brief Find the size of the frame in the file. (256x256, 256x1024, 512x1024)
+     * @param fname path to the file
+     * @throws std::runtime_error if the file is empty or the size cannot be determined
+     */
+     void find_frame_size(const std::filesystem::path &fname);
+
+
+     void parse_fname(const std::filesystem::path &fname);
+     void scan_files();
+     void open_file(size_t file_index);
+     std::filesystem::path fpath(size_t frame_index) const;
+
+
+  };
+
+} // namespace aare

include/aare/NDArray.hpp

@@ -69,6 +69,11 @@ class NDArray : public ArrayExpr<NDArray<T, Ndim>, Ndim> {
         std::copy(v.begin(), v.end(), begin());
     }
 
+    template<size_t Size>
+    NDArray(const std::array<T, Size>& arr) : NDArray<T,1>({Size}) {
+        std::copy(arr.begin(), arr.end(), begin());
+    }
+
     // Move constructor
     NDArray(NDArray &&other) noexcept
         : shape_(other.shape_), strides_(c_strides<Ndim>(shape_)),
@@ -97,6 +102,9 @@ class NDArray : public ArrayExpr<NDArray<T, Ndim>, Ndim> {
     auto begin() { return data_; }
     auto end() { return data_ + size_; }
 
+    auto begin() const { return data_; }
+    auto end() const { return data_ + size_; }
+
     using value_type = T;
 
     NDArray &operator=(NDArray &&other) noexcept; // Move assign
@@ -105,6 +113,20 @@ class NDArray : public ArrayExpr<NDArray<T, Ndim>, Ndim> {
     NDArray &operator-=(const NDArray &other);
     NDArray &operator*=(const NDArray &other);
 
+    //Write directly to the data array, or create a new one
+    template<size_t Size>
+    NDArray<T,1>& operator=(const std::array<T,Size> &other){
+        if(Size != size_){
+            delete[] data_;
+            size_ = Size;
+            data_ = new T[size_];
+        }
+        for (size_t i = 0; i < Size; ++i) {
+            data_[i] = other[i];
+        }
+        return *this;
+    }
+
     // NDArray& operator/=(const NDArray& other);
 
     template <typename V> NDArray &operator/=(const NDArray<V, Ndim> &other) {
@@ -135,6 +157,11 @@ class NDArray : public ArrayExpr<NDArray<T, Ndim>, Ndim> {
 
     NDArray &operator&=(const T & /*mask*/);
 
+  
+
+
+    
+
     void sqrt() {
         for (int i = 0; i < size_; ++i) {
             data_[i] = std::sqrt(data_[i]);
@@ -167,7 +194,7 @@ class NDArray : public ArrayExpr<NDArray<T, Ndim>, Ndim> {
 
     T *data() { return data_; }
     std::byte *buffer() { return reinterpret_cast<std::byte *>(data_); }
-    size_t size() const { return size_; }
+    ssize_t size() const { return static_cast<ssize_t>(size_); }
     size_t total_bytes() const { return size_ * sizeof(T); }
     std::array<int64_t, Ndim> shape() const noexcept { return shape_; }
     int64_t shape(int64_t i) const noexcept { return shape_[i]; }
@@ -318,6 +345,9 @@ NDArray<T, Ndim> &NDArray<T, Ndim>::operator+=(const T &value) {
     return *this;
 }
 
+
+
+
 template <typename T, int64_t Ndim>
 NDArray<T, Ndim> NDArray<T, Ndim>::operator+(const T &value) {
     NDArray result = *this;
@@ -361,12 +391,12 @@ NDArray<T, Ndim> NDArray<T, Ndim>::operator*(const T &value) {
     result *= value;
     return result;
 }
-template <typename T, int64_t Ndim> void NDArray<T, Ndim>::Print() {
-    if (shape_[0] < 20 && shape_[1] < 20)
-        Print_all();
-    else
-        Print_some();
-}
+// template <typename T, int64_t Ndim> void NDArray<T, Ndim>::Print() {
+//     if (shape_[0] < 20 && shape_[1] < 20)
+//         Print_all();
+//     else
+//         Print_some();
+// }
 
 template <typename T, int64_t Ndim>
 std::ostream &operator<<(std::ostream &os, const NDArray<T, Ndim> &arr) {
@@ -418,4 +448,6 @@ NDArray<T, Ndim> load(const std::string &pathname,
     return img;
 }
 
+
+
 } // namespace aare

include/aare/NDView.hpp

@@ -1,5 +1,5 @@
 #pragma once
-
+#include "aare/defs.hpp"
 #include "aare/ArrayExpr.hpp"
 
 #include <algorithm>
@@ -71,7 +71,7 @@ template <typename T, int64_t Ndim = 2> class NDView : public ArrayExpr<NDView<T
         return buffer_[element_offset(strides_, index...)];
     }
 
-    size_t size() const { return size_; }
+    ssize_t size() const { return static_cast<ssize_t>(size_); }
     size_t total_bytes() const { return size_ * sizeof(T); }
     std::array<int64_t, Ndim> strides() const noexcept { return strides_; }
 
@@ -99,6 +99,15 @@ template <typename T, int64_t Ndim = 2> class NDView : public ArrayExpr<NDView<T
 
     NDView &operator/=(const NDView &other) { return elemenwise(other, std::divides<T>()); }
 
+
+    template<size_t Size>
+    NDView& operator=(const std::array<T, Size> &arr) {
+        if(size() != static_cast<ssize_t>(arr.size()))
+            throw std::runtime_error(LOCATION + "Array and NDView size mismatch");
+        std::copy(arr.begin(), arr.end(), begin());
+        return *this;
+    }
+
     NDView &operator=(const T val) {
         for (auto it = begin(); it != end(); ++it)
             *it = val;
@@ -175,4 +184,9 @@ std::ostream& operator <<(std::ostream& os, const NDView<T, Ndim>& arr){
 }
 
 
+template <typename T>
+NDView<T,1> make_view(std::vector<T>& vec){
+    return NDView<T,1>(vec.data(), {static_cast<int64_t>(vec.size())});
+}
+
 } // namespace aare

include/aare/RawSubFile.hpp

@@ -22,7 +22,7 @@ class RawSubFile {
     size_t m_rows{};
     size_t m_cols{};
     size_t m_bytes_per_frame{};
-    size_t n_frames{};
+    size_t m_num_frames{};
     uint32_t m_pos_row{};
     uint32_t m_pos_col{};
  
@@ -53,6 +53,7 @@ class RawSubFile {
     size_t tell();
 
     void read_into(std::byte *image_buf, DetectorHeader *header = nullptr);
+    void read_into(std::byte *image_buf, size_t n_frames, DetectorHeader *header= nullptr);
     void get_part(std::byte *buffer, size_t frame_index);
     
     void read_header(DetectorHeader *header);
@@ -64,7 +65,9 @@ class RawSubFile {
 
     size_t bytes_per_frame() const { return m_bytes_per_frame; }
     size_t pixels_per_frame() const { return m_rows * m_cols; }
-    size_t bytes_per_pixel() const { return m_bitdepth / 8; }
+    size_t bytes_per_pixel() const { return m_bitdepth / bits_per_byte; }
+
+    size_t frames_in_file() const { return m_num_frames; }
 
 private:
   template <typename T>

include/aare/VarClusterFinder.hpp

@@ -7,7 +7,7 @@
 
 #include "aare/NDArray.hpp"
 
-const int MAX_CLUSTER_SIZE = 200;
+const int MAX_CLUSTER_SIZE = 50;
 namespace aare {
 
 template <typename T> class VarClusterFinder {
@@ -226,7 +226,7 @@ template <typename T> void VarClusterFinder<T>::single_pass(NDView<T, 2> img) {
 
 template <typename T> void VarClusterFinder<T>::first_pass() {
 
-    for (size_t i = 0; i < original_.size(); ++i) {
+    for (ssize_t i = 0; i < original_.size(); ++i) {
         if (use_noise_map)
             threshold_ = 5 * noiseMap(i);
         binary_(i) = (original_(i) > threshold_);
@@ -250,7 +250,7 @@ template <typename T> void VarClusterFinder<T>::first_pass() {
 
 template <typename T> void VarClusterFinder<T>::second_pass() {
 
-    for (size_t i = 0; i != labeled_.size(); ++i) {
+    for (ssize_t i = 0; i != labeled_.size(); ++i) {
         auto cl = labeled_(i);
         if (cl != 0) {
             auto it = child.find(cl);

include/aare/algorithm.hpp

@@ -0,0 +1,111 @@
+
+#pragma once
+#include <algorithm>
+#include <array>
+#include <vector>
+#include <aare/NDArray.hpp>
+
+namespace aare {
+/**
+ * @brief Index of the last element that is smaller than val.
+ * Requires a sorted array. Uses >= for ordering. If all elements
+ * are smaller it returns the last element and if all elements are
+ * larger it returns the first element.
+ * @param first iterator to the first element
+ * @param last iterator to the last element
+ * @param val value to compare
+ * @return index of the last element that is smaller than val
+ * 
+ */
+template <typename T>
+size_t last_smaller(const T* first, const T* last, T val) {
+    for (auto iter = first+1; iter != last; ++iter) {
+        if (*iter >= val) {
+            return std::distance(first, iter-1);
+        }
+    }
+    return std::distance(first, last-1);
+}
+
+template <typename T>
+size_t last_smaller(const NDArray<T, 1>& arr, T val) {
+    return last_smaller(arr.begin(), arr.end(), val);
+}
+
+template <typename T>
+size_t last_smaller(const std::vector<T>& vec, T val) {
+    return last_smaller(vec.data(), vec.data()+vec.size(), val);
+}
+
+/**
+ * @brief Index of the first element that is larger than val.
+ * Requires a sorted array. Uses > for ordering. If all elements
+ * are larger it returns the first element and if all elements are
+ * smaller it returns the last element.
+ * @param first iterator to the first element
+ * @param last iterator to the last element
+ * @param val value to compare
+ * @return index of the first element that is larger than val
+ */
+template <typename T>
+size_t first_larger(const T* first, const T* last, T val) {
+    for (auto iter = first; iter != last; ++iter) {
+        if (*iter > val) {
+            return std::distance(first, iter);
+        }
+    }
+    return std::distance(first, last-1);
+}
+
+template <typename T>
+size_t first_larger(const NDArray<T, 1>& arr, T val) {
+    return first_larger(arr.begin(), arr.end(), val);
+}
+
+template <typename T>
+size_t first_larger(const std::vector<T>& vec, T val) {
+    return first_larger(vec.data(), vec.data()+vec.size(), val);
+}
+
+/**
+ * @brief Index of the nearest element to val.
+ * Requires a sorted array. If there is no difference it takes the first element.
+ * @param first iterator to the first element
+ * @param last iterator to the last element
+ * @param val value to compare
+ * @return index of the nearest element
+ */
+template <typename T>
+size_t nearest_index(const T* first, const T* last, T val) {
+    auto iter = std::min_element(first, last,
+    [val](T a, T b) {
+        return std::abs(a - val) < std::abs(b - val);
+    });
+    return std::distance(first, iter);
+}
+
+template <typename T>
+size_t nearest_index(const NDArray<T, 1>& arr, T val) {
+    return nearest_index(arr.begin(), arr.end(), val);
+}
+
+template <typename T>
+size_t nearest_index(const std::vector<T>& vec, T val) {
+    return nearest_index(vec.data(), vec.data()+vec.size(), val);
+}
+
+template <typename T, size_t N>
+size_t nearest_index(const std::array<T,N>& arr, T val) {
+    return nearest_index(arr.data(), arr.data()+arr.size(), val);
+}
+
+template <typename T>
+std::vector<T> cumsum(const std::vector<T>& vec) {
+    std::vector<T> result(vec.size());
+    std::partial_sum(vec.begin(), vec.end(), result.begin());
+    return result;
+}
+
+
+
+} // namespace aare

include/aare/decode.hpp

@@ -1,6 +1,7 @@
 #pragma once
 
 #include <cstdint>
+#include <vector>
 #include <aare/NDView.hpp>
 namespace aare {
 
@@ -10,4 +11,16 @@ uint16_t adc_sar_04_decode64to16(uint64_t input);
 void adc_sar_05_decode64to16(NDView<uint64_t, 2> input, NDView<uint16_t,2> output);
 void adc_sar_04_decode64to16(NDView<uint64_t, 2> input, NDView<uint16_t,2> output);
 
-} // namespace aare
+
+/**
+ * @brief Apply custom weights to a 16-bit input value. Will sum up weights[i]**i
+ * for each bit i that is set in the input value.
+ * @throws std::out_of_range if weights.size() < 16
+ * @param input 16-bit input value
+ * @param weights vector of weights, size must be less than or equal to 16
+ */
+double apply_custom_weights(uint16_t input, const NDView<double, 1> weights);
+
+void apply_custom_weights(NDView<uint16_t, 1> input, NDView<double, 1> output, const NDView<double, 1> weights);
+
+} // namespace aare

include/aare/defs.hpp

@@ -1,11 +1,9 @@
 #pragma once
 
 #include "aare/Dtype.hpp"
-// #include "aare/utils/logger.hpp"
 
 #include <array>
 #include <stdexcept>
-
 #include <cassert>
 #include <cstdint>
 #include <cstring>
@@ -38,9 +36,12 @@
 
 namespace aare {
 
+inline constexpr size_t bits_per_byte = 8;
+
 void assert_failed(const std::string &msg);
 
 
+
 class DynamicCluster {
   public:
     int cluster_sizeX;
@@ -213,6 +214,9 @@ struct ROI{
   
     int64_t height() const { return ymax - ymin; }
     int64_t width() const { return xmax - xmin; }
+    bool contains(int64_t x, int64_t y) const {
+        return x >= xmin && x < xmax && y >= ymin && y < ymax;
+    }
   };
 
 

include/aare/utils/ifstream_helpers.hpp

@@ -0,0 +1,12 @@
+#pragma once
+
+#include <fstream>
+#include <string>
+namespace aare {
+
+/** 
+ * @brief Get the error message from an ifstream object
+*/
+std::string ifstream_error_msg(std::ifstream &ifs);
+
+} // namespace aare

include/aare/utils/par.hpp

@@ -0,0 +1,18 @@
+#include <thread>
+#include <vector>
+#include <utility>
+
+namespace aare {
+
+    template<typename F>
+    void RunInParallel(F func, const std::vector<std::pair<int, int>>& tasks) {
+        // auto tasks = split_task(0, y.shape(0), n_threads);
+        std::vector<std::thread> threads;
+        for (auto &task : tasks) {
+            threads.push_back(std::thread(func, task.first, task.second));
+        }
+        for (auto &thread : threads) {
+            thread.join();
+        }
+    }
+} // namespace aare

patches/libzmq_cmake_version.patch

@@ -0,0 +1,18 @@
+diff --git a/CMakeLists.txt b/CMakeLists.txt
+index dd3d8eb9..c0187747 100644
+--- a/CMakeLists.txt
++++ b/CMakeLists.txt
+@@ -1,11 +1,8 @@
+ # CMake build script for ZeroMQ
+ project(ZeroMQ)
+ 
+-if(${CMAKE_SYSTEM_NAME} STREQUAL Darwin)
+-  cmake_minimum_required(VERSION 3.0.2)
+-else()
+-  cmake_minimum_required(VERSION 2.8.12)
+-endif()
++cmake_minimum_required(VERSION 3.15)
++message(STATUS "Patched cmake version")
+ 
+ include(CheckIncludeFiles)
+ include(CheckCCompilerFlag)

pyproject.toml

@@ -4,12 +4,32 @@ build-backend = "scikit_build_core.build"
 
 [project]
 name = "aare"
-version = "2025.2.12"
+version = "2025.4.22"
+requires-python = ">=3.11"
+dependencies = [
+    "numpy",
+    "matplotlib",
+]
+
+
+[tool.cibuildwheel]
+
+build = "cp{311,312,313}-manylinux_x86_64"
+
+
+
 
 [tool.scikit-build]
-cmake.verbose = true
+build.verbose = true
+cmake.build-type = "Release"
+install.components = ["python"]
 
 [tool.scikit-build.cmake.define]
 AARE_PYTHON_BINDINGS = "ON"
-AARE_SYSTEM_LIBRARIES = "ON"
-AARE_INSTALL_PYTHONEXT = "ON"
+AARE_INSTALL_PYTHONEXT = "ON"
+
+
+[tool.pytest.ini_options]
+markers = [
+    "files: marks tests that need additional data (deselect with '-m \"not files\"')",
+]

python/CMakeLists.txt

@@ -1,12 +1,13 @@
 
-find_package (Python 3.10 COMPONENTS Interpreter Development REQUIRED)
+find_package (Python 3.10 COMPONENTS Interpreter Development.Module REQUIRED)
+set(PYBIND11_FINDPYTHON ON) # Needed for RH8
 
 # Download or find pybind11 depending on configuration
 if(AARE_FETCH_PYBIND11)
     FetchContent_Declare(
         pybind11
         GIT_REPOSITORY https://github.com/pybind/pybind11
-        GIT_TAG        v2.13.0
+        GIT_TAG        v2.13.6
     )
     FetchContent_MakeAvailable(pybind11)
 else()
@@ -28,6 +29,9 @@ target_link_libraries(_aare PRIVATE aare_core aare_compiler_flags)
 set( PYTHON_FILES
     aare/__init__.py
     aare/CtbRawFile.py
+    aare/ClusterFinder.py
+    aare/ClusterVector.py
+
     aare/func.py
     aare/RawFile.py
     aare/transform.py
@@ -35,6 +39,7 @@ set( PYTHON_FILES
     aare/utils.py
 )
 
+
 # Copy the python files to the build directory
 foreach(FILE ${PYTHON_FILES})
     configure_file(${FILE} ${CMAKE_BINARY_DIR}/${FILE}  )
@@ -50,18 +55,24 @@ set(PYTHON_EXAMPLES
 )
 
 
-
 # Copy the python examples to the build directory
 foreach(FILE ${PYTHON_EXAMPLES})
     configure_file(${FILE} ${CMAKE_BINARY_DIR}/${FILE}  )
+    message(STATUS "Copying ${FILE} to ${CMAKE_BINARY_DIR}/${FILE}")
 endforeach(FILE ${PYTHON_EXAMPLES})
 
 
 if(AARE_INSTALL_PYTHONEXT)
-    install(TARGETS _aare
+    install(
+        TARGETS _aare
         EXPORT "${TARGETS_EXPORT_NAME}"
         LIBRARY DESTINATION aare
+        COMPONENT python
     )
 
-    install(FILES ${PYTHON_FILES} DESTINATION aare)
+    install(
+        FILES ${PYTHON_FILES} 
+        DESTINATION aare
+        COMPONENT python
+        )
 endif()

python/aare/ClusterFinder.py

@@ -0,0 +1,67 @@
+
+from ._aare import ClusterFinder_Cluster3x3i, ClusterFinder_Cluster2x2i, ClusterFinderMT_Cluster3x3i, ClusterFinderMT_Cluster2x2i, ClusterCollector_Cluster3x3i, ClusterCollector_Cluster2x2i
+
+
+from ._aare import ClusterFileSink_Cluster3x3i, ClusterFileSink_Cluster2x2i
+import numpy as np
+
+def ClusterFinder(image_size, cluster_size, n_sigma=5, dtype = np.int32, capacity = 1024):
+    """
+    Factory function to create a ClusterFinder object. Provides a cleaner syntax for 
+    the templated ClusterFinder in C++.
+    """
+    if dtype == np.int32 and cluster_size == (3,3):
+        return ClusterFinder_Cluster3x3i(image_size, n_sigma = n_sigma, capacity=capacity)
+    elif dtype == np.int32 and cluster_size == (2,2): 
+        return ClusterFinder_Cluster2x2i(image_size, n_sigma = n_sigma, capacity=capacity)
+    else:
+        #TODO! add the other formats
+        raise ValueError(f"Unsupported dtype: {dtype}. Only np.int32 is supported.")
+
+
+def ClusterFinderMT(image_size, cluster_size = (3,3), dtype=np.int32, n_sigma=5, capacity = 1024, n_threads = 3): 
+    """ 
+    Factory function to create a ClusterFinderMT object. Provides a cleaner syntax for 
+    the templated ClusterFinderMT in C++.
+    """
+
+    if dtype == np.int32 and cluster_size == (3,3):
+        return ClusterFinderMT_Cluster3x3i(image_size, n_sigma = n_sigma,
+                    capacity = capacity, n_threads = n_threads)
+    elif dtype == np.int32 and cluster_size == (2,2):
+        return ClusterFinderMT_Cluster2x2i(image_size, n_sigma = n_sigma,
+                    capacity = capacity, n_threads = n_threads)
+    else:
+        #TODO! add the other formats
+        raise ValueError(f"Unsupported dtype: {dtype}. Only np.int32 is supported.")
+
+
+def ClusterCollector(clusterfindermt, cluster_size = (3,3), dtype=np.int32): 
+    """ 
+    Factory function to create a ClusterCollector object. Provides a cleaner syntax for 
+    the templated ClusterCollector in C++.
+    """
+
+    if dtype == np.int32 and cluster_size == (3,3): 
+        return ClusterCollector_Cluster3x3i(clusterfindermt)
+    elif dtype == np.int32 and cluster_size == (2,2):
+        return ClusterCollector_Cluster2x2i(clusterfindermt)
+     
+    else: 
+        #TODO! add the other formats
+        raise ValueError(f"Unsupported dtype: {dtype}. Only np.int32 is supported.")
+
+def ClusterFileSink(clusterfindermt, cluster_file, dtype=np.int32): 
+    """ 
+    Factory function to create a ClusterCollector object. Provides a cleaner syntax for 
+    the templated ClusterCollector in C++.
+    """
+
+    if dtype == np.int32 and clusterfindermt.cluster_size == (3,3): 
+        return ClusterFileSink_Cluster3x3i(clusterfindermt, cluster_file)
+    elif dtype == np.int32 and clusterfindermt.cluster_size == (2,2):
+        return ClusterFileSink_Cluster2x2i(clusterfindermt, cluster_file)
+     
+    else: 
+        #TODO! add the other formats
+        raise ValueError(f"Unsupported dtype: {dtype}. Only np.int32 is supported.")

python/aare/ClusterVector.py

@@ -0,0 +1,11 @@
+
+
+from ._aare import ClusterVector_Cluster3x3i
+import numpy as np
+
+def ClusterVector(cluster_size, dtype = np.int32):
+
+    if dtype == np.int32 and cluster_size == (3,3):
+        return ClusterVector_Cluster3x3i()
+    else:
+        raise ValueError(f"Unsupported dtype: {dtype}. Only np.int32 is supported.")

python/aare/__init__.py

@@ -2,21 +2,31 @@
 from . import _aare
 
 
-from ._aare import File, RawMasterFile, RawSubFile
-from ._aare import Pedestal_d, Pedestal_f, ClusterFinder, VarClusterFinder
+from ._aare import File, RawMasterFile, RawSubFile, JungfrauDataFile
+from ._aare import Pedestal_d, Pedestal_f, ClusterFinder_Cluster3x3i, VarClusterFinder
 from ._aare import DetectorType
-from ._aare import ClusterFile
+from ._aare import ClusterFile_Cluster3x3i as ClusterFile
 from ._aare import hitmap
+from ._aare import ROI
+
+# from ._aare import ClusterFinderMT, ClusterCollector, ClusterFileSink, ClusterVector_i
+
+from .ClusterFinder import ClusterFinder, ClusterCollector, ClusterFinderMT, ClusterFileSink
+from .ClusterVector import ClusterVector
 
-from ._aare import ClusterFinderMT, ClusterCollector, ClusterFileSink, ClusterVector_i
 
 from ._aare import fit_gaus, fit_pol1
+from ._aare import Interpolator
+from ._aare import calculate_eta2
+
+
+from ._aare import apply_custom_weights
 
 from .CtbRawFile import CtbRawFile
 from .RawFile import RawFile
 from .ScanParameters import ScanParameters
 
-from .utils import random_pixels, random_pixel, flat_list
+from .utils import random_pixels, random_pixel, flat_list, add_colorbar
 
 
 #make functions available in the top level API

python/aare/utils.py

@@ -1,4 +1,6 @@
 import numpy as np
+import matplotlib.pyplot as plt
+from mpl_toolkits.axes_grid1 import make_axes_locatable
 
 def random_pixels(n_pixels, xmin=0, xmax=512, ymin=0, ymax=1024):
     """Return a list of random pixels.
@@ -24,4 +26,11 @@ def random_pixel(xmin=0, xmax=512, ymin=0, ymax=1024):
 
 def flat_list(xss):
     """Flatten a list of lists."""
-    return [x for xs in xss for x in xs]
+    return [x for xs in xss for x in xs]
+
+def add_colorbar(ax, im, size="5%", pad=0.05):
+    """Add a colorbar with the same height as the image."""
+    divider = make_axes_locatable(ax)
+    cax = divider.append_axes("right", size=size, pad=pad)
+    plt.colorbar(im, cax=cax)
+    return ax, im, cax

python/examples/play.py

@@ -1,68 +1,79 @@
 import sys
 sys.path.append('/home/l_msdetect/erik/aare/build')
 
-#Our normal python imports
-from pathlib import Path
-import matplotlib.pyplot as plt
+from aare._aare import ClusterVector_i, Interpolator
+
+import pickle 
 import numpy as np
+import matplotlib.pyplot as plt
 import boost_histogram as bh
+import torch
+import math
 import time
 
-<<<<<<< HEAD
-from aare import File, ClusterFinder, VarClusterFinder, ClusterFile, CtbRawFile
-from aare import gaus, fit_gaus
 
-base = Path('/mnt/sls_det_storage/moench_data/Julian/MOENCH05/20250113_first_xrays_redo/raw_files/')
-cluster_file = Path('/home/l_msdetect/erik/tmp/Cu.clust')
 
-t0 = time.perf_counter()
-offset= -0.5
-hist3d = bh.Histogram(
-    bh.axis.Regular(160, 0+offset, 160+offset),  #x
-    bh.axis.Regular(150, 0+offset, 150+offset),  #y
-    bh.axis.Regular(200, 0, 6000), #ADU
-)
+def gaussian_2d(mx, my, sigma = 1, res=100, grid_size = 2):
+    """
+    Generate a 2D gaussian as position mx, my, with sigma=sigma. 
+    The gaussian is placed on a 2x2 pixel matrix with resolution 
+    res in one dimesion.
+    """
+    x = torch.linspace(0, pixel_size*grid_size, res)
+    x,y = torch.meshgrid(x,x, indexing="ij")
+    return 1 / (2*math.pi*sigma**2) * \
+      torch.exp(-((x - my)**2 / (2*sigma**2) + (y - mx)**2 / (2*sigma**2)))
 
-total_clusters = 0
-with ClusterFile(cluster_file, chunk_size = 1000) as f:
-    for i, clusters in enumerate(f):
-        arr = np.array(clusters)
-        total_clusters += clusters.size
-        hist3d.fill(arr['y'],arr['x'], clusters.sum_2x2()) #python talks [row, col] cluster finder [x,y]
-=======
-from aare import RawFile
+scale = 1000 #Scale factor when converting to integer
+pixel_size = 25 #um
+grid = 2
+resolution = 100
+sigma_um = 10
+xa = np.linspace(0,grid*pixel_size,resolution)
+ticks = [0, 25, 50]
 
-f = RawFile('/mnt/sls_det_storage/jungfrau_data1/vadym_tests/jf12_M431/laser_scan/laserScan_pedestal_G0_master_0.json')
+hit = np.array((20,20))
+etahist_fname = "/home/l_msdetect/erik/tmp/test_hist.pkl"
 
-print(f'{f.frame_number(1)}')
+local_resolution = 99
+grid_size = 3
+xaxis = np.linspace(0,grid_size*pixel_size, local_resolution)
+t = gaussian_2d(hit[0],hit[1], grid_size = grid_size, sigma = 10, res = local_resolution)
+pixels = t.reshape(grid_size, t.shape[0] // grid_size, grid_size, t.shape[1] // grid_size).sum(axis = 3).sum(axis = 1)
+pixels = pixels.numpy()
+pixels = (pixels*scale).astype(np.int32)
+v = ClusterVector_i(3,3)
+v.push_back(1,1, pixels)
 
-for i in range(10):
-    header, img = f.read_frame()
-    print(header['frameNumber'], img.shape)
->>>>>>> developer
+with open(etahist_fname, "rb") as f:
+        hist = pickle.load(f)
+eta = hist.view().copy()
+etabinsx = np.array(hist.axes.edges.T[0].flat)
+etabinsy = np.array(hist.axes.edges.T[1].flat)
+ebins = np.array(hist.axes.edges.T[2].flat)
+p = Interpolator(eta, etabinsx[0:-1], etabinsy[0:-1], ebins[0:-1])
 
-        
-t_elapsed = time.perf_counter()-t0
-print(f'Histogram filling took: {t_elapsed:.3f}s {total_clusters/t_elapsed/1e6:.3f}M clusters/s')
 
-histogram_data = hist3d.counts()
-x = hist3d.axes[2].edges[:-1]
 
-y = histogram_data[100,100,:]
-xx = np.linspace(x[0], x[-1])
-# fig, ax = plt.subplots()
-# ax.step(x, y, where = 'post')
 
-y_err = np.sqrt(y)
-y_err = np.zeros(y.size)
-y_err += 1
+#Generate the hit
 
-# par = fit_gaus2(y,x, y_err)
-# ax.plot(xx, gaus(xx,par))
-# print(par)
 
-res = fit_gaus(y,x)
-res2 = fit_gaus(y,x, y_err)
-print(res)
-print(res2)
 
+
+tmp = p.interpolate(v)
+print(f'tmp:{tmp}')
+pos = np.array((tmp['x'], tmp['y']))*25
+
+
+print(pixels)
+fig, ax = plt.subplots(figsize = (7,7))
+ax.pcolormesh(xaxis, xaxis, t)
+ax.plot(*pos, 'o')
+ax.set_xticks([0,25,50,75])
+ax.set_yticks([0,25,50,75])
+ax.set_xlim(0,75)
+ax.set_ylim(0,75)
+ax.grid()
+print(f'{hit=}')
+print(f'{pos=}')

python/src/bind_ClusterVector.hpp

@@ -0,0 +1,104 @@
+#include "aare/ClusterCollector.hpp"
+#include "aare/ClusterFileSink.hpp"
+#include "aare/ClusterFinder.hpp"
+#include "aare/ClusterFinderMT.hpp"
+#include "aare/ClusterVector.hpp"
+#include "aare/NDView.hpp"
+#include "aare/Pedestal.hpp"
+#include "np_helper.hpp"
+
+#include <cstdint>
+#include <filesystem>
+#include <pybind11/pybind11.h>
+#include <pybind11/stl.h>
+#include <pybind11/stl_bind.h>
+
+namespace py = pybind11;
+using pd_type = double;
+
+using namespace aare;
+
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+
+template <typename Type, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType = uint16_t>
+void define_ClusterVector(py::module &m, const std::string &typestr) {
+    using ClusterType = Cluster<Type, ClusterSizeX, ClusterSizeY, CoordType>;
+    auto class_name = fmt::format("ClusterVector_{}", typestr);
+
+    py::class_<ClusterVector<
+        Cluster<Type, ClusterSizeX, ClusterSizeY, CoordType>, void>>(
+        m, class_name.c_str(),
+        py::buffer_protocol())
+
+        .def(py::init()) // TODO change!!!
+
+        .def("push_back",
+             [](ClusterVector<ClusterType> &self, const ClusterType &cluster) {
+                 self.push_back(cluster);
+             })
+
+        .def("sum",
+             [](ClusterVector<ClusterType> &self) {
+                 auto *vec = new std::vector<Type>(self.sum());
+                 return return_vector(vec);
+             })
+        .def("sum_2x2", [](ClusterVector<ClusterType> &self){
+            auto *vec = new std::vector<Type>(self.sum_2x2());
+            return return_vector(vec);
+        })
+        .def_property_readonly("size", &ClusterVector<ClusterType>::size)
+        .def("item_size", &ClusterVector<ClusterType>::item_size)
+        .def_property_readonly("fmt",
+                               [typestr](ClusterVector<ClusterType> &self) {
+                                   return fmt_format<ClusterType>;
+                               })
+
+        .def_property_readonly("cluster_size_x",
+                               &ClusterVector<ClusterType>::cluster_size_x)
+        .def_property_readonly("cluster_size_y",
+                               &ClusterVector<ClusterType>::cluster_size_y)
+        .def_property_readonly("capacity",
+                               &ClusterVector<ClusterType>::capacity)
+        .def_property("frame_number", &ClusterVector<ClusterType>::frame_number,
+                      &ClusterVector<ClusterType>::set_frame_number)
+        .def_buffer(
+            [typestr](ClusterVector<ClusterType> &self) -> py::buffer_info {
+                return py::buffer_info(
+                    self.data(),             /* Pointer to buffer */
+                    self.item_size(),        /* Size of one scalar */
+                    fmt_format<ClusterType>, /* Format descriptor */
+                    1,                       /* Number of dimensions */
+                    {self.size()},           /* Buffer dimensions */
+                    {self.item_size()} /* Strides (in bytes) for each index */
+                );
+            });
+
+    // Free functions using ClusterVector
+    m.def("hitmap",
+          [](std::array<size_t, 2> image_size, ClusterVector<ClusterType> &cv) {
+              // Create a numpy array to hold the hitmap
+              // The shape of the array is (image_size[0], image_size[1])
+              // note that the python array is passed as [row, col] which
+              // is the opposite of the clusters [x,y]
+              py::array_t<int32_t> hitmap(image_size);
+              auto r = hitmap.mutable_unchecked<2>();
+
+              // Initialize hitmap to 0
+              for (py::ssize_t i = 0; i < r.shape(0); i++)
+                  for (py::ssize_t j = 0; j < r.shape(1); j++)
+                      r(i, j) = 0;
+
+              // Loop over the clusters and increment the hitmap
+              // Skip out of bound clusters
+              for (const auto &cluster : cv) {
+                  auto x = cluster.x;
+                  auto y = cluster.y;
+                  if (x < image_size[1] && y < image_size[0])
+                      r(cluster.y, cluster.x) += 1;
+              }
+
+              return hitmap;
+          });
+}

python/src/cluster.hpp

@@ -16,186 +16,196 @@
 namespace py = pybind11;
 using pd_type = double;
 
-template <typename T>
-void define_cluster_vector(py::module &m, const std::string &typestr) {
-    auto class_name = fmt::format("ClusterVector_{}", typestr);
-    py::class_<ClusterVector<T>>(m, class_name.c_str(), py::buffer_protocol())
-        .def(py::init<int, int>())
-        .def_property_readonly("size", &ClusterVector<T>::size)
-        .def("item_size", &ClusterVector<T>::item_size)
-        .def_property_readonly("fmt",
-                               [typestr](ClusterVector<T> &self) {
-                                   return fmt::format(
-                                       self.fmt_base(), self.cluster_size_x(),
-                                       self.cluster_size_y(), typestr);
-                               })
-        .def("sum",
-             [](ClusterVector<T> &self) {
-                 auto *vec = new std::vector<T>(self.sum());
-                 return return_vector(vec);
-             })
-        .def("sum_2x2", [](ClusterVector<T> &self) {
-            auto *vec = new std::vector<T>(self.sum_2x2());
-            return return_vector(vec);
-        })
-        .def_property_readonly("capacity", &ClusterVector<T>::capacity)
-        .def_property("frame_number", &ClusterVector<T>::frame_number,
-                      &ClusterVector<T>::set_frame_number)
-        .def_buffer([typestr](ClusterVector<T> &self) -> py::buffer_info {
-            return py::buffer_info(
-                self.data(),      /* Pointer to buffer */
-                self.item_size(), /* Size of one scalar */
-                fmt::format(self.fmt_base(), self.cluster_size_x(),
-                            self.cluster_size_y(),
-                            typestr), /* Format descriptor */
-                1,                    /* Number of dimensions */
-                {self.size()},        /* Buffer dimensions */
-                {self.item_size()}    /* Strides (in bytes) for each index */
-            );
+using namespace aare;
+
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+
+template <typename Type, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType>
+void define_cluster(py::module &m, const std::string &typestr) {
+    auto class_name = fmt::format("Cluster{}", typestr);
+
+    py::class_<Cluster<Type, ClusterSizeX, ClusterSizeY, CoordType>>(
+        m, class_name.c_str(), py::buffer_protocol())
+
+        .def(py::init([](uint8_t x, uint8_t y, py::array_t<Type> data) {
+            py::buffer_info buf_info = data.request();
+            Cluster<Type, ClusterSizeX, ClusterSizeY, CoordType> cluster;
+            cluster.x = x;
+            cluster.y = y;
+            auto r = data.template unchecked<1>(); // no bounds checks
+            for (py::ssize_t i = 0; i < data.size(); ++i) {
+                cluster.data[i] = r(i);
+            }
+            return cluster;
+        }));
+
+    /*
+    .def_property(
+        "data",
+        [](ClusterType &c) -> py::array {
+            return py::array(py::buffer_info(
+                c.data, sizeof(Type),
+                py::format_descriptor<Type>::format(), // Type
+                                                       // format
+                1, // Number of dimensions
+                {static_cast<ssize_t>(ClusterSizeX *
+                                      ClusterSizeY)}, // Shape (flattened)
+                {sizeof(Type)} // Stride (step size between elements)
+                ));
+        },
+        [](ClusterType &c, py::array_t<Type> arr) {
+            py::buffer_info buf_info = arr.request();
+            Type *ptr = static_cast<Type *>(buf_info.ptr);
+            std::copy(ptr, ptr + ClusterSizeX * ClusterSizeY,
+                      c.data); // TODO dont iterate over centers!!!
+
         });
+    */
 }
 
-void define_cluster_finder_mt_bindings(py::module &m) {
-    py::class_<ClusterFinderMT<uint16_t, pd_type>>(m, "ClusterFinderMT")
-        .def(py::init<Shape<2>, Shape<2>, pd_type, size_t, size_t>(),
-             py::arg("image_size"), py::arg("cluster_size"),
-             py::arg("n_sigma") = 5.0, py::arg("capacity") = 2048,
-             py::arg("n_threads") = 3)
+template <typename T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType = uint16_t>
+void define_cluster_finder_mt_bindings(py::module &m,
+                                       const std::string &typestr) {
+    auto class_name = fmt::format("ClusterFinderMT_{}", typestr);
+
+    using ClusterType = Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>;
+
+    py::class_<ClusterFinderMT<ClusterType, uint16_t, pd_type>>(
+        m, class_name.c_str())
+        .def(py::init<Shape<2>, pd_type, size_t, size_t>(),
+             py::arg("image_size"), py::arg("n_sigma") = 5.0,
+             py::arg("capacity") = 2048, py::arg("n_threads") = 3)
         .def("push_pedestal_frame",
-             [](ClusterFinderMT<uint16_t, pd_type> &self,
+             [](ClusterFinderMT<ClusterType, uint16_t, pd_type> &self,
                 py::array_t<uint16_t> frame) {
                  auto view = make_view_2d(frame);
                  self.push_pedestal_frame(view);
              })
         .def(
             "find_clusters",
-            [](ClusterFinderMT<uint16_t, pd_type> &self,
+            [](ClusterFinderMT<ClusterType, uint16_t, pd_type> &self,
                py::array_t<uint16_t> frame, uint64_t frame_number) {
                 auto view = make_view_2d(frame);
                 self.find_clusters(view, frame_number);
                 return;
             },
             py::arg(), py::arg("frame_number") = 0)
-        .def("clear_pedestal", &ClusterFinderMT<uint16_t, pd_type>::clear_pedestal)
-        .def("sync", &ClusterFinderMT<uint16_t, pd_type>::sync)
-        .def("stop", &ClusterFinderMT<uint16_t, pd_type>::stop)
-        .def("start", &ClusterFinderMT<uint16_t, pd_type>::start)
-        .def("pedestal",
-             [](ClusterFinderMT<uint16_t, pd_type> &self, size_t thread_index) {
-                 auto pd = new NDArray<pd_type, 2>{};
-                 *pd = self.pedestal(thread_index);
-                 return return_image_data(pd);
-             },py::arg("thread_index") = 0)
-        .def("noise",
-             [](ClusterFinderMT<uint16_t, pd_type> &self, size_t thread_index) {
-                 auto arr = new NDArray<pd_type, 2>{};
-                 *arr = self.noise(thread_index);
-                 return return_image_data(arr);
-             },py::arg("thread_index") = 0);
+        .def_property_readonly("cluster_size", [](ClusterFinderMT<ClusterType, uint16_t, pd_type> &self){
+            return py::make_tuple(ClusterSizeX, ClusterSizeY);
+        })
+        .def("clear_pedestal",
+             &ClusterFinderMT<ClusterType, uint16_t, pd_type>::clear_pedestal)
+        .def("sync", &ClusterFinderMT<ClusterType, uint16_t, pd_type>::sync)
+        .def("stop", &ClusterFinderMT<ClusterType, uint16_t, pd_type>::stop)
+        .def("start", &ClusterFinderMT<ClusterType, uint16_t, pd_type>::start)
+        .def(
+            "pedestal",
+            [](ClusterFinderMT<ClusterType, uint16_t, pd_type> &self,
+               size_t thread_index) {
+                auto pd = new NDArray<pd_type, 2>{};
+                *pd = self.pedestal(thread_index);
+                return return_image_data(pd);
+            },
+            py::arg("thread_index") = 0)
+        .def(
+            "noise",
+            [](ClusterFinderMT<ClusterType, uint16_t, pd_type> &self,
+               size_t thread_index) {
+                auto arr = new NDArray<pd_type, 2>{};
+                *arr = self.noise(thread_index);
+                return return_image_data(arr);
+            },
+            py::arg("thread_index") = 0);
 }
 
-void define_cluster_collector_bindings(py::module &m) {
-    py::class_<ClusterCollector>(m, "ClusterCollector")
-        .def(py::init<ClusterFinderMT<uint16_t, double, int32_t> *>())
-        .def("stop", &ClusterCollector::stop)
+template <typename T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType = uint16_t>
+void define_cluster_collector_bindings(py::module &m,
+                                       const std::string &typestr) {
+    auto class_name = fmt::format("ClusterCollector_{}", typestr);
+
+    using ClusterType = Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>;
+
+    py::class_<ClusterCollector<ClusterType>>(m, class_name.c_str())
+        .def(py::init<ClusterFinderMT<ClusterType, uint16_t, double> *>())
+        .def("stop", &ClusterCollector<ClusterType>::stop)
         .def(
             "steal_clusters",
-            [](ClusterCollector &self) {
-                auto v =
-                    new std::vector<ClusterVector<int>>(self.steal_clusters());
-                return v;
+            [](ClusterCollector<ClusterType> &self) {
+                auto v = new std::vector<ClusterVector<ClusterType>>(
+                    self.steal_clusters());
+                return v; // TODO change!!!
             },
             py::return_value_policy::take_ownership);
 }
 
-void define_cluster_file_sink_bindings(py::module &m) {
-    py::class_<ClusterFileSink>(m, "ClusterFileSink")
-        .def(py::init<ClusterFinderMT<uint16_t, double, int32_t> *,
+template <typename T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType = uint16_t>
+void define_cluster_file_sink_bindings(py::module &m,
+                                       const std::string &typestr) {
+    auto class_name = fmt::format("ClusterFileSink_{}", typestr);
+
+    using ClusterType = Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>;
+
+    py::class_<ClusterFileSink<ClusterType>>(m, class_name.c_str())
+        .def(py::init<ClusterFinderMT<ClusterType, uint16_t, double> *,
                       const std::filesystem::path &>())
-        .def("stop", &ClusterFileSink::stop);
+        .def("stop", &ClusterFileSink<ClusterType>::stop);
 }
 
-void define_cluster_finder_bindings(py::module &m) {
-    py::class_<ClusterFinder<uint16_t, pd_type>>(m, "ClusterFinder")
-        .def(py::init<Shape<2>, Shape<2>, pd_type, size_t>(),
-             py::arg("image_size"), py::arg("cluster_size"),
+template <typename T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType = uint16_t>
+void define_cluster_finder_bindings(py::module &m, const std::string &typestr) {
+    auto class_name = fmt::format("ClusterFinder_{}", typestr);
+
+    using ClusterType = Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>;
+
+    py::class_<ClusterFinder<ClusterType, uint16_t, pd_type>>(
+        m, class_name.c_str())
+        .def(py::init<Shape<2>, pd_type, size_t>(), py::arg("image_size"),
              py::arg("n_sigma") = 5.0, py::arg("capacity") = 1'000'000)
         .def("push_pedestal_frame",
-             [](ClusterFinder<uint16_t, pd_type> &self,
+             [](ClusterFinder<ClusterType, uint16_t, pd_type> &self,
                 py::array_t<uint16_t> frame) {
                  auto view = make_view_2d(frame);
                  self.push_pedestal_frame(view);
              })
-        .def("clear_pedestal", &ClusterFinder<uint16_t, pd_type>::clear_pedestal)
-        .def_property_readonly("pedestal",
-                               [](ClusterFinder<uint16_t, pd_type> &self) {
-                                   auto pd = new NDArray<pd_type, 2>{};
-                                   *pd = self.pedestal();
-                                   return return_image_data(pd);
-                               })
-        .def_property_readonly("noise",
-                               [](ClusterFinder<uint16_t, pd_type> &self) {
-                                   auto arr = new NDArray<pd_type, 2>{};
-                                   *arr = self.noise();
-                                   return return_image_data(arr);
-                               })
+        .def("clear_pedestal",
+             &ClusterFinder<ClusterType, uint16_t, pd_type>::clear_pedestal)
+        .def_property_readonly(
+            "pedestal",
+            [](ClusterFinder<ClusterType, uint16_t, pd_type> &self) {
+                auto pd = new NDArray<pd_type, 2>{};
+                *pd = self.pedestal();
+                return return_image_data(pd);
+            })
+        .def_property_readonly(
+            "noise",
+            [](ClusterFinder<ClusterType, uint16_t, pd_type> &self) {
+                auto arr = new NDArray<pd_type, 2>{};
+                *arr = self.noise();
+                return return_image_data(arr);
+            })
         .def(
             "steal_clusters",
-            [](ClusterFinder<uint16_t, pd_type> &self,
+            [](ClusterFinder<ClusterType, uint16_t, pd_type> &self,
                bool realloc_same_capacity) {
-                auto v = new ClusterVector<int>(
-                    self.steal_clusters(realloc_same_capacity));
-                return v;
+                ClusterVector<ClusterType> clusters =
+                    self.steal_clusters(realloc_same_capacity);
+                return clusters;
             },
             py::arg("realloc_same_capacity") = false)
         .def(
             "find_clusters",
-            [](ClusterFinder<uint16_t, pd_type> &self,
+            [](ClusterFinder<ClusterType, uint16_t, pd_type> &self,
                py::array_t<uint16_t> frame, uint64_t frame_number) {
                 auto view = make_view_2d(frame);
                 self.find_clusters(view, frame_number);
                 return;
             },
             py::arg(), py::arg("frame_number") = 0);
-
-    m.def("hitmap",
-          [](std::array<size_t, 2> image_size, ClusterVector<int32_t> &cv) {
-              py::array_t<int32_t> hitmap(image_size);
-              auto r = hitmap.mutable_unchecked<2>();
-
-              // Initialize hitmap to 0
-              for (py::ssize_t i = 0; i < r.shape(0); i++)
-                  for (py::ssize_t j = 0; j < r.shape(1); j++)
-                      r(i, j) = 0;
-
-              size_t stride = cv.item_size();
-              auto ptr = cv.data();
-              for (size_t i = 0; i < cv.size(); i++) {
-                  auto x = *reinterpret_cast<int16_t *>(ptr);
-                  auto y = *reinterpret_cast<int16_t *>(ptr + sizeof(int16_t));
-                  r(y, x) += 1;
-                  ptr += stride;
-              }
-              return hitmap;
-          });
-    define_cluster_vector<int>(m, "i");
-    define_cluster_vector<double>(m, "d");
-    define_cluster_vector<float>(m, "f");
-
-    py::class_<DynamicCluster>(m, "DynamicCluster", py::buffer_protocol())
-        .def(py::init<int, int, Dtype>())
-        .def("size", &DynamicCluster::size)
-        .def("begin", &DynamicCluster::begin)
-        .def("end", &DynamicCluster::end)
-        .def_readwrite("x", &DynamicCluster::x)
-        .def_readwrite("y", &DynamicCluster::y)
-        .def_buffer([](DynamicCluster &c) -> py::buffer_info {
-            return py::buffer_info(c.data(), c.dt.bytes(), c.dt.format_descr(),
-                                   1, {c.size()}, {c.dt.bytes()});
-        })
-
-        .def("__repr__", [](const DynamicCluster &a) {
-            return "<DynamicCluster: x: " + std::to_string(a.x) +
-                   ", y: " + std::to_string(a.y) + ">";
-        });
-}
+}
+#pragma GCC diagnostic pop

python/src/cluster_file.hpp

@@ -1,3 +1,4 @@
+#include "aare/CalculateEta.hpp"
 #include "aare/ClusterFile.hpp"
 #include "aare/defs.hpp"
 
@@ -10,63 +11,84 @@
 #include <pybind11/stl/filesystem.h>
 #include <string>
 
-//Disable warnings for unused parameters, as we ignore some
-//in the __exit__ method
+// Disable warnings for unused parameters, as we ignore some
+// in the __exit__ method
 #pragma GCC diagnostic push
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 
-
 namespace py = pybind11;
 using namespace ::aare;
 
-void define_cluster_file_io_bindings(py::module &m) {
-    PYBIND11_NUMPY_DTYPE(Cluster3x3, x, y, data);
+template <typename Type, uint8_t CoordSizeX, uint8_t CoordSizeY,
+          typename CoordType = uint16_t>
+void define_cluster_file_io_bindings(py::module &m,
+                                     const std::string &typestr) {
 
-    py::class_<ClusterFile>(m, "ClusterFile")
+    using ClusterType = Cluster<Type, CoordSizeX, CoordSizeY, CoordType>;
+
+    auto class_name = fmt::format("ClusterFile_{}", typestr);
+
+    py::class_<ClusterFile<ClusterType>>(m, class_name.c_str())
         .def(py::init<const std::filesystem::path &, size_t,
                       const std::string &>(),
              py::arg(), py::arg("chunk_size") = 1000, py::arg("mode") = "r")
-        .def("read_clusters",
-             [](ClusterFile &self, size_t n_clusters) {
-                auto v = new ClusterVector<int32_t>(self.read_clusters(n_clusters));
+        .def(
+            "read_clusters",
+            [](ClusterFile<ClusterType> &self, size_t n_clusters) {
+                auto v = new ClusterVector<ClusterType>(
+                    self.read_clusters(n_clusters));
                 return v;
-             },py::return_value_policy::take_ownership)
+            },
+            py::return_value_policy::take_ownership)
         .def("read_frame",
-             [](ClusterFile &self) {
-                auto v = new ClusterVector<int32_t>(self.read_frame());
-                return v;
+             [](ClusterFile<ClusterType> &self) {
+                 auto v = new ClusterVector<ClusterType>(self.read_frame());
+                 return v;
              })
-        .def("write_frame", &ClusterFile::write_frame)
-        // .def("read_cluster_with_cut",
-        //      [](ClusterFile &self, size_t n_clusters,
-        //         py::array_t<double> noise_map, int nx, int ny) {
-        //          auto view = make_view_2d(noise_map);
-        //          auto *vec =
-        //              new std::vector<Cluster3x3>(self.read_cluster_with_cut(
-        //                  n_clusters, view.data(), nx, ny));
-        //          return return_vector(vec);
-        //      })
-        .def("__enter__", [](ClusterFile &self) { return &self; })
+        .def("set_roi", &ClusterFile<ClusterType>::set_roi)
+        .def(
+            "set_noise_map",
+            [](ClusterFile<ClusterType> &self, py::array_t<int32_t> noise_map) {
+                auto view = make_view_2d(noise_map);
+                self.set_noise_map(view);
+            })
+
+        .def("set_gain_map",
+             [](ClusterFile<ClusterType> &self, py::array_t<double> gain_map) {
+                 auto view = make_view_2d(gain_map);
+                 self.set_gain_map(view);
+             })
+
+        .def("close", &ClusterFile<ClusterType>::close)
+        .def("write_frame", &ClusterFile<ClusterType>::write_frame)
+        .def("__enter__", [](ClusterFile<ClusterType> &self) { return &self; })
         .def("__exit__",
-             [](ClusterFile &self,
+             [](ClusterFile<ClusterType> &self,
                 const std::optional<pybind11::type> &exc_type,
                 const std::optional<pybind11::object> &exc_value,
                 const std::optional<pybind11::object> &traceback) {
                  self.close();
              })
-        .def("__iter__", [](ClusterFile &self) { return &self; })
-        .def("__next__", [](ClusterFile &self) {
-            auto v = new ClusterVector<int32_t>(self.read_clusters(self.chunk_size()));
+        .def("__iter__", [](ClusterFile<ClusterType> &self) { return &self; })
+        .def("__next__", [](ClusterFile<ClusterType> &self) {
+            auto v = new ClusterVector<ClusterType>(
+                self.read_clusters(self.chunk_size()));
             if (v->size() == 0) {
                 throw py::stop_iteration();
             }
             return v;
         });
+}
 
-        m.def("calculate_eta2", []( aare::ClusterVector<int32_t> &clusters) {
-            auto eta2 = new NDArray<double, 2>(calculate_eta2(clusters));
-            return return_image_data(eta2);
-        });
+template <typename Type, uint8_t CoordSizeX, uint8_t CoordSizeY,
+          typename CoordType = uint16_t>
+void register_calculate_eta(py::module &m) {
+    using ClusterType = Cluster<Type, CoordSizeX, CoordSizeY, CoordType>;
+    m.def("calculate_eta2",
+          [](const aare::ClusterVector<ClusterType> &clusters) {
+              auto eta2 = new NDArray<double, 2>(calculate_eta2(clusters));
+              return return_image_data(eta2);
+          });
 }
 
 #pragma GCC diagnostic pop

python/src/ctb_raw_file.hpp

@@ -10,6 +10,8 @@
 #include "aare/decode.hpp"
 // #include "aare/fClusterFileV2.hpp"
 
+#include "np_helper.hpp"
+
 #include <cstdint>
 #include <filesystem>
 #include <pybind11/iostream.h>
@@ -32,7 +34,7 @@ m.def("adc_sar_05_decode64to16", [](py::array_t<uint8_t> input) {
     }
 
     //Create a 2D output array with the same shape as the input
-    std::vector<ssize_t> shape{input.shape(0), input.shape(1)/8};
+    std::vector<ssize_t> shape{input.shape(0), input.shape(1)/static_cast<int64_t>(bits_per_byte)};
     py::array_t<uint16_t> output(shape);
 
     //Create a view of the input and output arrays
@@ -53,7 +55,7 @@ m.def("adc_sar_04_decode64to16", [](py::array_t<uint8_t> input) {
     }
 
     //Create a 2D output array with the same shape as the input
-    std::vector<ssize_t> shape{input.shape(0), input.shape(1)/8};
+    std::vector<ssize_t> shape{input.shape(0), input.shape(1)/static_cast<int64_t>(bits_per_byte)};
     py::array_t<uint16_t> output(shape);
 
     //Create a view of the input and output arrays
@@ -65,35 +67,54 @@ m.def("adc_sar_04_decode64to16", [](py::array_t<uint8_t> input) {
     return output;
 });
 
- py::class_<CtbRawFile>(m, "CtbRawFile")
-        .def(py::init<const std::filesystem::path &>())
-        .def("read_frame",
-             [](CtbRawFile &self) {
-                 size_t image_size = self.image_size_in_bytes();
-                 py::array image;
-                 std::vector<ssize_t> shape;
-                 shape.reserve(2);
-                 shape.push_back(1);
-                 shape.push_back(image_size);
+m.def(
+    "apply_custom_weights",
+    [](py::array_t<uint16_t, py::array::c_style | py::array::forcecast> &input,
+       py::array_t<double, py::array::c_style | py::array::forcecast>
+           &weights) {
+        
 
-                 py::array_t<DetectorHeader> header(1);
+        // Create new array with same shape as the input array (uninitialized values)
+        py::buffer_info buf = input.request();
+        py::array_t<double> output(buf.shape);
 
-                 // always read bytes
-                 image = py::array_t<uint8_t>(shape);
+        // Use NDViews to call into the C++ library
+        auto weights_view = make_view_1d(weights);
+        NDView<uint16_t, 1> input_view(input.mutable_data(), {input.size()});
+        NDView<double, 1> output_view(output.mutable_data(), {output.size()});
 
-                 self.read_into(
-                     reinterpret_cast<std::byte *>(image.mutable_data()),
-                     header.mutable_data());
+        apply_custom_weights(input_view, output_view, weights_view);
+        return output;
+    });
 
-                 return py::make_tuple(header, image);
-             })
-        .def("seek", &CtbRawFile::seek)
-        .def("tell", &CtbRawFile::tell)
-        .def("master", &CtbRawFile::master)
+py::class_<CtbRawFile>(m, "CtbRawFile")
+    .def(py::init<const std::filesystem::path &>())
+    .def("read_frame",
+         [](CtbRawFile &self) {
+             size_t image_size = self.image_size_in_bytes();
+             py::array image;
+             std::vector<ssize_t> shape;
+             shape.reserve(2);
+             shape.push_back(1);
+             shape.push_back(image_size);
 
-        .def_property_readonly("image_size_in_bytes",
-                               &CtbRawFile::image_size_in_bytes)
+             py::array_t<DetectorHeader> header(1);
 
-        .def_property_readonly("frames_in_file", &CtbRawFile::frames_in_file);
+             // always read bytes
+             image = py::array_t<uint8_t>(shape);
 
-}
+             self.read_into(reinterpret_cast<std::byte *>(image.mutable_data()),
+                            header.mutable_data());
+
+             return py::make_tuple(header, image);
+         })
+    .def("seek", &CtbRawFile::seek)
+    .def("tell", &CtbRawFile::tell)
+    .def("master", &CtbRawFile::master)
+
+    .def_property_readonly("image_size_in_bytes",
+                           &CtbRawFile::image_size_in_bytes)
+
+    .def_property_readonly("frames_in_file", &CtbRawFile::frames_in_file);
+
+}

python/src/file.hpp

@@ -20,6 +20,9 @@
 namespace py = pybind11;
 using namespace ::aare;
 
+
+
+
 //Disable warnings for unused parameters, as we ignore some
 //in the __exit__ method
 #pragma GCC diagnostic push
@@ -195,6 +198,8 @@ void define_file_io_bindings(py::module &m) {
 
     py::class_<ROI>(m, "ROI")
         .def(py::init<>())
+        .def(py::init<int64_t, int64_t, int64_t, int64_t>(), py::arg("xmin"),
+             py::arg("xmax"), py::arg("ymin"), py::arg("ymax"))
         .def_readwrite("xmin", &ROI::xmin)
         .def_readwrite("xmax", &ROI::xmax)
         .def_readwrite("ymin", &ROI::ymin)
@@ -212,36 +217,9 @@ void define_file_io_bindings(py::module &m) {
     
 
 
-    py::class_<RawSubFile>(m, "RawSubFile")
-        .def(py::init<const std::filesystem::path &, DetectorType, size_t,
-                      size_t, size_t>())
-        .def_property_readonly("bytes_per_frame", &RawSubFile::bytes_per_frame)
-        .def_property_readonly("pixels_per_frame",
-                               &RawSubFile::pixels_per_frame)
-        .def("seek", &RawSubFile::seek)
-        .def("tell", &RawSubFile::tell)
-        .def_property_readonly("rows", &RawSubFile::rows)
-        .def_property_readonly("cols", &RawSubFile::cols)
-        .def("read_frame",
-             [](RawSubFile &self) {
-                 const uint8_t item_size = self.bytes_per_pixel();
-                 py::array image;
-                 std::vector<ssize_t> shape;
-                 shape.reserve(2);
-                 shape.push_back(self.rows());
-                 shape.push_back(self.cols());
-                 if (item_size == 1) {
-                     image = py::array_t<uint8_t>(shape);
-                 } else if (item_size == 2) {
-                     image = py::array_t<uint16_t>(shape);
-                 } else if (item_size == 4) {
-                     image = py::array_t<uint32_t>(shape);
-                 }
-                 fmt::print("item_size: {} rows: {} cols: {}\n", item_size, self.rows(), self.cols());
-                 self.read_into(
-                     reinterpret_cast<std::byte *>(image.mutable_data()));
-                 return image;
-             });
+
+
+
 
 #pragma GCC diagnostic pop
     // py::class_<ClusterHeader>(m, "ClusterHeader")

python/src/fit.hpp

@@ -7,6 +7,8 @@
 #include "aare/Fit.hpp"
 
 namespace py = pybind11;
+using namespace pybind11::literals;
+
 
 void define_fit_bindings(py::module &m) {
 
@@ -29,7 +31,8 @@ void define_fit_bindings(py::module &m) {
             The points at which to evaluate the Gaussian function.
         par : array_like
             The parameters of the Gaussian function. The first element is the amplitude, the second element is the mean, and the third element is the standard deviation.
-        )", py::arg("x"), py::arg("par"));
+        )",
+        py::arg("x"), py::arg("par"));
 
     m.def(
         "pol1",
@@ -49,7 +52,9 @@ void define_fit_bindings(py::module &m) {
             The points at which to evaluate the polynomial function.
         par : array_like
             The parameters of the polynomial function. The first element is the intercept, and the second element is the slope.    
-        )", py::arg("x"), py::arg("par"));
+        )",
+        py::arg("x"), py::arg("par"));
+
 
     m.def(
         "fit_gaus",
@@ -72,7 +77,8 @@ void define_fit_bindings(py::module &m) {
                 throw std::runtime_error("Data must be 1D or 3D");
             }
         },
-R"(
+        R"(
+
 Fit a 1D Gaussian to data.
 
 Parameters
@@ -90,8 +96,9 @@ n_threads : int, optional
         "fit_gaus",
         [](py::array_t<double, py::array::c_style | py::array::forcecast> x,
            py::array_t<double, py::array::c_style | py::array::forcecast> y,
-           py::array_t<double, py::array::c_style | py::array::forcecast>
-               y_err, int n_threads) {
+           py::array_t<double, py::array::c_style | py::array::forcecast> y_err,
+           int n_threads) {
+
             if (y.ndim() == 3) {
                 // Allocate memory for the output
                 // Need to have pointers to allow python to manage
@@ -99,15 +106,20 @@ n_threads : int, optional
                 auto par = new NDArray<double, 3>({y.shape(0), y.shape(1), 3});
                 auto par_err =
                     new NDArray<double, 3>({y.shape(0), y.shape(1), 3});
+                auto chi2 = new NDArray<double, 2>({y.shape(0), y.shape(1)});
 
+                // Make views of the numpy arrays
                 auto y_view = make_view_3d(y);
                 auto y_view_err = make_view_3d(y_err);
                 auto x_view = make_view_1d(x);
+
                 aare::fit_gaus(x_view, y_view, y_view_err, par->view(),
-                               par_err->view(), n_threads);
-                // return return_image_data(par);
-                return py::make_tuple(return_image_data(par),
-                                      return_image_data(par_err));
+                               par_err->view(), chi2->view(), n_threads);
+
+                return py::dict("par"_a = return_image_data(par),
+                                "par_err"_a = return_image_data(par_err),
+                                "chi2"_a = return_image_data(chi2),
+                                "Ndf"_a = y.shape(2) - 3);
             } else if (y.ndim() == 1) {
                 // Allocate memory for the output
                 // Need to have pointers to allow python to manage
@@ -120,15 +132,21 @@ n_threads : int, optional
                 auto y_view_err = make_view_1d(y_err);
                 auto x_view = make_view_1d(x);
 
+
+                double chi2 = 0;
                 aare::fit_gaus(x_view, y_view, y_view_err, par->view(),
-                               par_err->view());
-                return py::make_tuple(return_image_data(par),
-                                      return_image_data(par_err));
+                               par_err->view(), chi2);
+
+                return py::dict("par"_a = return_image_data(par),
+                                "par_err"_a = return_image_data(par_err),
+                                "chi2"_a = chi2, "Ndf"_a = y.size() - 3);
+
             } else {
                 throw std::runtime_error("Data must be 1D or 3D");
             }
         },
-R"(
+        R"(
+
 Fit a 1D Gaussian to data with error estimates.
 
 Parameters
@@ -172,11 +190,11 @@ n_threads : int, optional
         "fit_pol1",
         [](py::array_t<double, py::array::c_style | py::array::forcecast> x,
            py::array_t<double, py::array::c_style | py::array::forcecast> y,
-           py::array_t<double, py::array::c_style | py::array::forcecast>
-               y_err, int n_threads) {
+           py::array_t<double, py::array::c_style | py::array::forcecast> y_err,
+           int n_threads) {
             if (y.ndim() == 3) {
-                auto par =
-                    new NDArray<double, 3>({y.shape(0), y.shape(1), 2});
+                auto par = new NDArray<double, 3>({y.shape(0), y.shape(1), 2});
+
                 auto par_err =
                     new NDArray<double, 3>({y.shape(0), y.shape(1), 2});
 
@@ -184,10 +202,15 @@ n_threads : int, optional
                 auto y_view_err = make_view_3d(y_err);
                 auto x_view = make_view_1d(x);
 
-                aare::fit_pol1(x_view, y_view,y_view_err, par->view(),
-                               par_err->view(), n_threads);
-                return py::make_tuple(return_image_data(par),
-                                      return_image_data(par_err));
+                auto chi2 = new NDArray<double, 2>({y.shape(0), y.shape(1)});
+
+                aare::fit_pol1(x_view, y_view, y_view_err, par->view(),
+                               par_err->view(), chi2->view(), n_threads);
+                return py::dict("par"_a = return_image_data(par),
+                                "par_err"_a = return_image_data(par_err), 
+                                "chi2"_a = return_image_data(chi2),
+                                "Ndf"_a = y.shape(2) - 2);
+
 
             } else if (y.ndim() == 1) {
                 auto par = new NDArray<double, 1>({2});
@@ -197,15 +220,19 @@ n_threads : int, optional
                 auto y_view_err = make_view_1d(y_err);
                 auto x_view = make_view_1d(x);
 
+                double chi2 = 0;
+
                 aare::fit_pol1(x_view, y_view, y_view_err, par->view(),
-                               par_err->view());
-                return py::make_tuple(return_image_data(par),
-                                      return_image_data(par_err));
+                               par_err->view(), chi2);
+                return py::dict("par"_a = return_image_data(par),
+                                "par_err"_a = return_image_data(par_err),
+                                "chi2"_a = chi2, "Ndf"_a = y.size() - 2);
+
             } else {
                 throw std::runtime_error("Data must be 1D or 3D");
             }
         },
-R"(
+        R"(
 Fit a 1D polynomial to data with error estimates.
 
 Parameters

python/src/interpolation.hpp

@@ -0,0 +1,82 @@
+#include "aare/Interpolator.hpp"
+#include "aare/NDArray.hpp"
+#include "aare/NDView.hpp"
+#include "np_helper.hpp"
+#include <cstdint>
+#include <filesystem>
+#include <pybind11/pybind11.h>
+#include <pybind11/stl.h>
+
+namespace py = pybind11;
+
+template <typename Type, uint8_t CoordSizeX, uint8_t CoordSizeY,
+          typename CoordType = uint16_t>
+void register_interpolate(py::class_<aare::Interpolator> &interpolator) {
+
+    using ClusterType = Cluster<Type, CoordSizeX, CoordSizeY, CoordType>;
+
+    interpolator.def("interpolate",
+                     [](aare::Interpolator &self,
+                        const ClusterVector<ClusterType> &clusters) {
+                         auto photons = self.interpolate<ClusterType>(clusters);
+                         auto *ptr = new std::vector<Photon>{photons};
+                         return return_vector(ptr);
+                     });
+}
+
+void define_interpolation_bindings(py::module &m) {
+
+    PYBIND11_NUMPY_DTYPE(aare::Photon, x, y, energy);
+
+    auto interpolator =
+        py::class_<aare::Interpolator>(m, "Interpolator")
+            .def(py::init([](py::array_t<double, py::array::c_style |
+                                                     py::array::forcecast>
+                                 etacube,
+                             py::array_t<double> xbins,
+                             py::array_t<double> ybins,
+                             py::array_t<double> ebins) {
+                return Interpolator(make_view_3d(etacube), make_view_1d(xbins),
+                                    make_view_1d(ybins), make_view_1d(ebins));
+            }))
+            .def("get_ietax",
+                 [](Interpolator &self) {
+                     auto *ptr = new NDArray<double, 3>{};
+                     *ptr = self.get_ietax();
+                     return return_image_data(ptr);
+                 })
+            .def("get_ietay", [](Interpolator &self) {
+                auto *ptr = new NDArray<double, 3>{};
+                *ptr = self.get_ietay();
+                return return_image_data(ptr);
+            });
+
+    register_interpolate<int, 3, 3, uint16_t>(interpolator);
+    register_interpolate<float, 3, 3, uint16_t>(interpolator);
+    register_interpolate<double, 3, 3, uint16_t>(interpolator);
+    register_interpolate<int, 2, 2, uint16_t>(interpolator);
+    register_interpolate<float, 2, 2, uint16_t>(interpolator);
+    register_interpolate<double, 2, 2, uint16_t>(interpolator);
+
+    // TODO! Evaluate without converting to double
+    m.def(
+        "hej",
+        []() {
+            // auto boost_histogram = py::module_::import("boost_histogram");
+            // py::object axis =
+            // boost_histogram.attr("axis").attr("Regular")(10, 0.0, 10.0);
+            // py::object histogram = boost_histogram.attr("Histogram")(axis);
+            // return histogram;
+            // return h;
+        },
+        R"(
+        Evaluate a 1D Gaussian function for all points in x using parameters par.
+
+        Parameters
+        ----------
+        x : array_like
+            The points at which to evaluate the Gaussian function.
+        par : array_like
+            The parameters of the Gaussian function. The first element is the amplitude, the second element is the mean, and the third element is the standard deviation.
+        )");
+}

python/src/jungfrau_data_file.hpp

@@ -0,0 +1,116 @@
+
+#include "aare/JungfrauDataFile.hpp"
+#include "aare/defs.hpp"
+
+#include <cstdint>
+#include <filesystem>
+#include <pybind11/iostream.h>
+#include <pybind11/numpy.h>
+#include <pybind11/pybind11.h>
+#include <pybind11/stl.h>
+#include <pybind11/stl/filesystem.h>
+#include <string>
+
+namespace py = pybind11;
+using namespace ::aare;
+
+// Disable warnings for unused parameters, as we ignore some
+// in the __exit__ method
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+
+auto read_dat_frame(JungfrauDataFile &self) {
+    py::array_t<JungfrauDataHeader> header(1);
+    py::array_t<uint16_t> image({
+        self.rows(), 
+        self.cols() 
+    });
+
+    self.read_into(reinterpret_cast<std::byte *>(image.mutable_data()),
+                   header.mutable_data());
+
+    return py::make_tuple(header, image);
+}
+
+auto read_n_dat_frames(JungfrauDataFile &self, size_t n_frames) {
+    // adjust for actual frames left in the file
+    n_frames = std::min(n_frames, self.total_frames() - self.tell());
+    if (n_frames == 0) {
+        throw std::runtime_error("No frames left in file");
+    }
+
+    py::array_t<JungfrauDataHeader> header(n_frames);
+    py::array_t<uint16_t> image({
+        n_frames, self.rows(), 
+        self.cols()});
+
+    self.read_into(reinterpret_cast<std::byte *>(image.mutable_data()),
+                   n_frames, header.mutable_data());
+
+    return py::make_tuple(header, image);
+}
+
+void define_jungfrau_data_file_io_bindings(py::module &m) {
+    // Make the JungfrauDataHeader usable from numpy
+    PYBIND11_NUMPY_DTYPE(JungfrauDataHeader, framenum, bunchid);
+
+    py::class_<JungfrauDataFile>(m, "JungfrauDataFile")
+        .def(py::init<const std::filesystem::path &>())
+        .def("seek", &JungfrauDataFile::seek,
+             R"(
+               Seek to the given frame index.
+               )")
+        .def("tell", &JungfrauDataFile::tell,
+             R"(
+               Get the current frame index.
+               )")
+        .def_property_readonly("rows", &JungfrauDataFile::rows)
+        .def_property_readonly("cols", &JungfrauDataFile::cols)
+        .def_property_readonly("base_name", &JungfrauDataFile::base_name)
+        .def_property_readonly("bytes_per_frame",
+                               &JungfrauDataFile::bytes_per_frame)
+        .def_property_readonly("pixels_per_frame",
+                               &JungfrauDataFile::pixels_per_frame)
+        .def_property_readonly("bytes_per_pixel",
+                               &JungfrauDataFile::bytes_per_pixel)
+        .def_property_readonly("bitdepth", &JungfrauDataFile::bitdepth)
+        .def_property_readonly("current_file", &JungfrauDataFile::current_file)
+        .def_property_readonly("total_frames", &JungfrauDataFile::total_frames)
+        .def_property_readonly("n_files", &JungfrauDataFile::n_files)
+        .def("read_frame", &read_dat_frame,
+             R"(
+               Read a single frame from the file.
+               )")
+        .def("read_n", &read_n_dat_frames,
+             R"(
+               Read maximum n_frames frames from the file.
+               )")
+        .def(
+            "read",
+            [](JungfrauDataFile &self) {
+                self.seek(0);
+                auto n_frames = self.total_frames();
+                return read_n_dat_frames(self, n_frames);
+            },
+            R"(
+              Read all frames from the file. Seeks to the beginning before reading.
+              )")
+        .def("__enter__", [](JungfrauDataFile &self) { return &self; })
+        .def("__exit__",
+             [](JungfrauDataFile &self,
+                const std::optional<pybind11::type> &exc_type,
+                const std::optional<pybind11::object> &exc_value,
+                const std::optional<pybind11::object> &traceback) {
+                 //  self.close();
+             })
+        .def("__iter__", [](JungfrauDataFile &self) { return &self; })
+        .def("__next__", [](JungfrauDataFile &self) {
+            try {
+                return read_dat_frame(self);
+            } catch (std::runtime_error &e) {
+                throw py::stop_iteration();
+            }
+        });
+}
+
+#pragma GCC diagnostic pop

python/src/module.cpp

@@ -1,16 +1,24 @@
-//Files with bindings to the different classes
-#include "file.hpp"
-#include "raw_file.hpp"
-#include "ctb_raw_file.hpp"
-#include "raw_master_file.hpp"
-#include "var_cluster.hpp"
-#include "pixel_map.hpp"
-#include "pedestal.hpp"
+// Files with bindings to the different classes
+
+//New style file naming
+#include "bind_ClusterVector.hpp"
+
+//TODO! migrate the other names
 #include "cluster.hpp"
 #include "cluster_file.hpp"
+#include "ctb_raw_file.hpp"
+#include "file.hpp"
 #include "fit.hpp"
+#include "interpolation.hpp"
+#include "raw_sub_file.hpp"
+#include "raw_master_file.hpp"
+#include "raw_file.hpp"
+#include "pixel_map.hpp"
+#include "var_cluster.hpp"
+#include "pedestal.hpp"
+#include "jungfrau_data_file.hpp"
 
-//Pybind stuff
+// Pybind stuff
 #include <pybind11/pybind11.h>
 #include <pybind11/stl.h>
 
@@ -19,17 +27,70 @@ namespace py = pybind11;
 PYBIND11_MODULE(_aare, m) {
     define_file_io_bindings(m);
     define_raw_file_io_bindings(m);
+    define_raw_sub_file_io_bindings(m);
     define_ctb_raw_file_io_bindings(m);
     define_raw_master_file_bindings(m);
     define_var_cluster_finder_bindings(m);
     define_pixel_map_bindings(m);
     define_pedestal_bindings<double>(m, "Pedestal_d");
     define_pedestal_bindings<float>(m, "Pedestal_f");
-    define_cluster_finder_bindings(m);
-    define_cluster_finder_mt_bindings(m);
-    define_cluster_file_io_bindings(m);
-    define_cluster_collector_bindings(m);
-    define_cluster_file_sink_bindings(m);
     define_fit_bindings(m);
+    define_interpolation_bindings(m);
+    define_jungfrau_data_file_io_bindings(m);
 
-}
+    define_cluster_file_io_bindings<int, 3, 3, uint16_t>(m, "Cluster3x3i");
+    define_cluster_file_io_bindings<double, 3, 3, uint16_t>(m, "Cluster3x3d");
+    define_cluster_file_io_bindings<float, 3, 3, uint16_t>(m, "Cluster3x3f");
+    define_cluster_file_io_bindings<int, 2, 2, uint16_t>(m, "Cluster2x2i");
+    define_cluster_file_io_bindings<float, 2, 2, uint16_t>(m, "Cluster2x2f");
+    define_cluster_file_io_bindings<double, 2, 2, uint16_t>(m, "Cluster2x2d");
+
+    define_ClusterVector<int, 3, 3, uint16_t>(m, "Cluster3x3i");
+    define_ClusterVector<double, 3, 3, uint16_t>(m, "Cluster3x3d");
+    define_ClusterVector<float, 3, 3, uint16_t>(m, "Cluster3x3f");
+    define_ClusterVector<int, 2, 2, uint16_t>(m, "Cluster2x2i");
+    define_ClusterVector<double, 2, 2, uint16_t>(m, "Cluster2x2d");
+    define_ClusterVector<float, 2, 2, uint16_t>(m, "Cluster2x2f");
+
+    define_cluster_finder_bindings<int, 3, 3, uint16_t>(m, "Cluster3x3i");
+    define_cluster_finder_bindings<double, 3, 3, uint16_t>(m, "Cluster3x3d");
+    define_cluster_finder_bindings<float, 3, 3, uint16_t>(m, "Cluster3x3f");
+    define_cluster_finder_bindings<int, 2, 2, uint16_t>(m, "Cluster2x2i");
+    define_cluster_finder_bindings<double, 2, 2, uint16_t>(m, "Cluster2x2d");
+    define_cluster_finder_bindings<float, 2, 2, uint16_t>(m, "Cluster2x2f");
+
+    define_cluster_finder_mt_bindings<int, 3, 3, uint16_t>(m, "Cluster3x3i");
+    define_cluster_finder_mt_bindings<double, 3, 3, uint16_t>(m, "Cluster3x3d");
+    define_cluster_finder_mt_bindings<float, 3, 3, uint16_t>(m, "Cluster3x3f");
+    define_cluster_finder_mt_bindings<int, 2, 2, uint16_t>(m, "Cluster2x2i");
+    define_cluster_finder_mt_bindings<double, 2, 2, uint16_t>(m, "Cluster2x2d");
+    define_cluster_finder_mt_bindings<float, 2, 2, uint16_t>(m, "Cluster2x2f");
+
+    define_cluster_file_sink_bindings<int, 3, 3, uint16_t>(m, "Cluster3x3i");
+    define_cluster_file_sink_bindings<double, 3, 3, uint16_t>(m, "Cluster3x3d");
+    define_cluster_file_sink_bindings<float, 3, 3, uint16_t>(m, "Cluster3x3f");
+    define_cluster_file_sink_bindings<int, 2, 2, uint16_t>(m, "Cluster2x2i");
+    define_cluster_file_sink_bindings<double, 2, 2, uint16_t>(m, "Cluster2x2d");
+    define_cluster_file_sink_bindings<float, 2, 2, uint16_t>(m, "Cluster2x2f");
+
+    define_cluster_collector_bindings<int, 3, 3, uint16_t>(m, "Cluster3x3i");
+    define_cluster_collector_bindings<double, 3, 3, uint16_t>(m, "Cluster3x3f");
+    define_cluster_collector_bindings<float, 3, 3, uint16_t>(m, "Cluster3x3d");
+    define_cluster_collector_bindings<int, 2, 2, uint16_t>(m, "Cluster2x2i");
+    define_cluster_collector_bindings<double, 2, 2, uint16_t>(m, "Cluster2x2f");
+    define_cluster_collector_bindings<float, 2, 2, uint16_t>(m, "Cluster2x2d");
+
+    define_cluster<int, 3, 3, uint16_t>(m, "3x3i");
+    define_cluster<float, 3, 3, uint16_t>(m, "3x3f");
+    define_cluster<double, 3, 3, uint16_t>(m, "3x3d");
+    define_cluster<int, 2, 2, uint16_t>(m, "2x2i");
+    define_cluster<float, 2, 2, uint16_t>(m, "2x2f");
+    define_cluster<double, 2, 2, uint16_t>(m, "2x2d");
+
+    register_calculate_eta<int, 3, 3, uint16_t>(m);
+    register_calculate_eta<float, 3, 3, uint16_t>(m);
+    register_calculate_eta<double, 3, 3, uint16_t>(m);
+    register_calculate_eta<int, 2, 2, uint16_t>(m);
+    register_calculate_eta<float, 2, 2, uint16_t>(m);
+    register_calculate_eta<double, 2, 2, uint16_t>(m);
+}

python/src/np_helper.hpp

@@ -10,6 +10,7 @@
 #include "aare/NDView.hpp"
 
 namespace py = pybind11;
+using namespace aare;
 
 // Pass image data back to python as a numpy array
 template <typename T, int64_t Ndim>
@@ -40,25 +41,46 @@ template <typename T> py::array return_vector(std::vector<T> *vec) {
 }
 
 // todo rewrite generic
-template <class T, int Flags> auto get_shape_3d(py::array_t<T, Flags> arr) {
+template <class T, int Flags>
+auto get_shape_3d(const py::array_t<T, Flags> &arr) {
     return aare::Shape<3>{arr.shape(0), arr.shape(1), arr.shape(2)};
 }
 
-template <class T, int Flags> auto make_view_3d(py::array_t<T, Flags> arr) {
+template <class T, int Flags> auto make_view_3d(py::array_t<T, Flags> &arr) {
     return aare::NDView<T, 3>(arr.mutable_data(), get_shape_3d<T, Flags>(arr));
 }
 
-template <class T, int Flags> auto get_shape_2d(py::array_t<T, Flags> arr) {
+template <class T, int Flags>
+auto get_shape_2d(const py::array_t<T, Flags> &arr) {
     return aare::Shape<2>{arr.shape(0), arr.shape(1)};
 }
 
-template <class T, int Flags> auto get_shape_1d(py::array_t<T, Flags> arr) {
+template <class T, int Flags>
+auto get_shape_1d(const py::array_t<T, Flags> &arr) {
     return aare::Shape<1>{arr.shape(0)};
 }
 
-template <class T, int Flags> auto make_view_2d(py::array_t<T, Flags> arr) {
+template <class T, int Flags> auto make_view_2d(py::array_t<T, Flags> &arr) {
     return aare::NDView<T, 2>(arr.mutable_data(), get_shape_2d<T, Flags>(arr));
 }
-template <class T, int Flags> auto make_view_1d(py::array_t<T, Flags> arr) {
+template <class T, int Flags> auto make_view_1d(py::array_t<T, Flags> &arr) {
     return aare::NDView<T, 1>(arr.mutable_data(), get_shape_1d<T, Flags>(arr));
-}
+}
+
+template <typename ClusterType> struct fmt_format_trait; // forward declaration
+
+template <typename T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType>
+struct fmt_format_trait<Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>> {
+
+    static std::string value() {
+        return fmt::format("T{{{}:x:{}:y:{}:data:}}",
+                           py::format_descriptor<CoordType>::format(),
+                           py::format_descriptor<CoordType>::format(),
+                           fmt::format("({},{}){}", ClusterSizeX, ClusterSizeY,
+                                       py::format_descriptor<T>::format()));
+    }
+};
+
+template <typename ClusterType>
+auto fmt_format = fmt_format_trait<ClusterType>::value();

python/src/raw_sub_file.hpp

@@ -0,0 +1,110 @@
+#include "aare/CtbRawFile.hpp"
+#include "aare/File.hpp"
+#include "aare/Frame.hpp"
+#include "aare/RawFile.hpp"
+#include "aare/RawMasterFile.hpp"
+#include "aare/RawSubFile.hpp"
+
+#include "aare/defs.hpp"
+// #include "aare/fClusterFileV2.hpp"
+
+#include <cstdint>
+#include <filesystem>
+#include <pybind11/iostream.h>
+#include <pybind11/numpy.h>
+#include <pybind11/pybind11.h>
+#include <pybind11/stl.h>
+#include <pybind11/stl/filesystem.h>
+#include <string>
+
+namespace py = pybind11;
+using namespace ::aare;
+
+auto read_frame_from_RawSubFile(RawSubFile &self) {
+    py::array_t<DetectorHeader> header(1);
+    const uint8_t item_size = self.bytes_per_pixel();
+    std::vector<ssize_t> shape{static_cast<ssize_t>(self.rows()),
+        static_cast<ssize_t>(self.cols())};
+    
+    py::array image;
+    if (item_size == 1) {
+        image = py::array_t<uint8_t>(shape);
+    } else if (item_size == 2) {
+        image = py::array_t<uint16_t>(shape);
+    } else if (item_size == 4) {
+        image = py::array_t<uint32_t>(shape);
+    }
+    self.read_into(reinterpret_cast<std::byte *>(image.mutable_data()),
+                   header.mutable_data());
+
+    return py::make_tuple(header, image);
+}
+
+auto read_n_frames_from_RawSubFile(RawSubFile &self, size_t n_frames) {
+    py::array_t<DetectorHeader> header(n_frames);
+    const uint8_t item_size = self.bytes_per_pixel();
+    std::vector<ssize_t> shape{
+        static_cast<ssize_t>(n_frames), 
+        static_cast<ssize_t>(self.rows()),
+        static_cast<ssize_t>(self.cols())
+    };
+    
+    py::array image;
+    if (item_size == 1) {
+        image = py::array_t<uint8_t>(shape);
+    } else if (item_size == 2) {
+        image = py::array_t<uint16_t>(shape);
+    } else if (item_size == 4) {
+        image = py::array_t<uint32_t>(shape);
+    }
+    self.read_into(reinterpret_cast<std::byte *>(image.mutable_data()), n_frames,
+                   header.mutable_data());
+
+    return py::make_tuple(header, image);
+}
+
+
+//Disable warnings for unused parameters, as we ignore some
+//in the __exit__ method
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+
+void define_raw_sub_file_io_bindings(py::module &m) {
+    py::class_<RawSubFile>(m, "RawSubFile")
+        .def(py::init<const std::filesystem::path &, DetectorType, size_t,
+                      size_t, size_t>())
+        .def_property_readonly("bytes_per_frame", &RawSubFile::bytes_per_frame)
+        .def_property_readonly("pixels_per_frame",
+                               &RawSubFile::pixels_per_frame)
+        .def_property_readonly("bytes_per_pixel", &RawSubFile::bytes_per_pixel)                    
+        .def("seek", &RawSubFile::seek)
+        .def("tell", &RawSubFile::tell)
+        .def_property_readonly("rows", &RawSubFile::rows)
+        .def_property_readonly("cols", &RawSubFile::cols)
+        .def_property_readonly("frames_in_file", &RawSubFile::frames_in_file)
+        .def("read_frame", &read_frame_from_RawSubFile)
+        .def("read_n", &read_n_frames_from_RawSubFile)
+        .def("read", [](RawSubFile &self){
+            self.seek(0);
+            auto n_frames = self.frames_in_file();
+            return read_n_frames_from_RawSubFile(self, n_frames);
+        })
+        .def("__enter__", [](RawSubFile &self) { return &self; })
+        .def("__exit__",
+             [](RawSubFile &self,
+                const std::optional<pybind11::type> &exc_type,
+                const std::optional<pybind11::object> &exc_value,
+                const std::optional<pybind11::object> &traceback) {
+             })
+        .def("__iter__", [](RawSubFile &self) { return &self; })
+        .def("__next__", [](RawSubFile &self) {
+            try {
+                return read_frame_from_RawSubFile(self);
+            } catch (std::runtime_error &e) {
+                throw py::stop_iteration();
+            }
+        });
+
+}
+
+#pragma GCC diagnostic pop

python/src/var_cluster.hpp

@@ -19,15 +19,24 @@ using namespace::aare;
 
 void define_var_cluster_finder_bindings(py::module &m) {
     PYBIND11_NUMPY_DTYPE(VarClusterFinder<double>::Hit, size, row, col,
-                         reserved, energy, max);
+                         reserved, energy, max, rows, cols, enes);
 
     py::class_<VarClusterFinder<double>>(m, "VarClusterFinder")
         .def(py::init<Shape<2>, double>())
         .def("labeled",
              [](VarClusterFinder<double> &self) {
-                 auto ptr = new NDArray<int, 2>(self.labeled());
+                 auto *ptr = new NDArray<int, 2>(self.labeled());
                  return return_image_data(ptr);
              })
+        .def("set_noiseMap",
+            [](VarClusterFinder<double> &self,
+                py::array_t<double, py::array::c_style | py::array::forcecast>
+                    noise_map) {
+                auto noise_map_span = make_view_2d(noise_map);
+                self.set_noiseMap(noise_map_span);
+            })
+        .def("set_peripheralThresholdFactor",
+             &VarClusterFinder<double>::set_peripheralThresholdFactor)
         .def("find_clusters",
              [](VarClusterFinder<double> &self,
                 py::array_t<double, py::array::c_style | py::array::forcecast>
@@ -35,6 +44,30 @@ void define_var_cluster_finder_bindings(py::module &m) {
                  auto view = make_view_2d(img);
                  self.find_clusters(view);
              })
+        .def("find_clusters_X",
+             [](VarClusterFinder<double> &self,
+                py::array_t<double, py::array::c_style | py::array::forcecast>
+                    img) {
+                 auto img_span = make_view_2d(img);
+                 self.find_clusters_X(img_span);
+             })
+        .def("single_pass",
+             [](VarClusterFinder<double> &self,
+                py::array_t<double, py::array::c_style | py::array::forcecast>
+                    img) {
+                 auto img_span = make_view_2d(img);
+                 self.single_pass(img_span);
+             })
+        .def("hits",
+             [](VarClusterFinder<double> &self) {
+                 auto ptr = new std::vector<VarClusterFinder<double>::Hit>(
+                     self.steal_hits());
+                 return return_vector(ptr);
+             })
+        .def("clear_hits",
+             [](VarClusterFinder<double> &self) {
+                 self.clear_hits();
+             })
         .def("steal_hits",
              [](VarClusterFinder<double> &self) {
                  auto ptr = new std::vector<VarClusterFinder<double>::Hit>(

python/tests/conftest.py

@@ -0,0 +1,34 @@
+import os
+from pathlib import Path
+import pytest
+
+
+
+def pytest_addoption(parser):
+    parser.addoption(
+        "--files", action="store_true", default=False, help="run slow tests"
+    )
+
+
+def pytest_configure(config):
+    config.addinivalue_line("markers", "files: mark test as needing image files to run")
+
+
+def pytest_collection_modifyitems(config, items):
+    if config.getoption("--files"):
+        return
+    skip = pytest.mark.skip(reason="need --files option to run")
+    for item in items:
+        if "files" in item.keywords:
+            item.add_marker(skip)
+
+
+@pytest.fixture
+def test_data_path():
+    env_value = os.environ.get("AARE_TEST_DATA")
+    if not env_value:
+        raise RuntimeError("Environment variable AARE_TEST_DATA is not set or is empty")
+
+    return Path(env_value)
+
+

python/tests/test_Cluster.py

@@ -0,0 +1,110 @@
+import pytest 
+import numpy as np
+
+from aare import _aare #import the C++ module
+from conftest import test_data_path
+
+
+def test_cluster_vector_can_be_converted_to_numpy():
+    cv = _aare.ClusterVector_Cluster3x3i()
+    arr = np.array(cv, copy=False)
+    assert arr.shape == (0,)  # 4 for x, y, size, energy and 9 for the cluster data
+
+
+def test_ClusterVector(): 
+    """Test ClusterVector""" 
+
+    clustervector = _aare.ClusterVector_Cluster3x3i()
+    assert clustervector.cluster_size_x == 3
+    assert clustervector.cluster_size_y == 3
+    assert clustervector.item_size() == 4+9*4
+    assert clustervector.frame_number == 0
+    assert clustervector.size == 0
+
+    cluster = _aare.Cluster3x3i(0,0,np.ones(9, dtype=np.int32))
+    
+    clustervector.push_back(cluster) 
+    assert clustervector.size == 1
+
+    with pytest.raises(TypeError):  # Or use the appropriate exception type
+        clustervector.push_back(_aare.Cluster2x2i(0,0,np.ones(4, dtype=np.int32)))
+
+    with pytest.raises(TypeError): 
+        clustervector.push_back(_aare.Cluster3x3f(0,0,np.ones(9, dtype=np.float32))) 
+
+def test_Interpolator(): 
+    """Test Interpolator"""
+
+    ebins = np.linspace(0,10, 20, dtype=np.float64)
+    xbins = np.linspace(0, 5, 30, dtype=np.float64)
+    ybins = np.linspace(0, 5, 30, dtype=np.float64)
+
+    etacube = np.zeros(shape=[30, 30, 20], dtype=np.float64)
+    interpolator = _aare.Interpolator(etacube, xbins, ybins, ebins)
+
+    assert interpolator.get_ietax().shape == (30,30,20)
+    assert interpolator.get_ietay().shape == (30,30,20)
+    clustervector = _aare.ClusterVector_Cluster3x3i()
+    
+    cluster = _aare.Cluster3x3i(0,0, np.ones(9, dtype=np.int32))
+    clustervector.push_back(cluster) 
+
+    interpolated_photons = interpolator.interpolate(clustervector)
+
+    assert interpolated_photons.size == 1
+
+    assert interpolated_photons[0]["x"] == -1 
+    assert interpolated_photons[0]["y"] == -1
+    assert interpolated_photons[0]["energy"] == 4 #eta_sum = 4, dx, dy = -1,-1 m_ietax = 0, m_ietay = 0
+
+    clustervector = _aare.ClusterVector_Cluster2x2i()
+
+    cluster = _aare.Cluster2x2i(0,0, np.ones(4, dtype=np.int32))
+    clustervector.push_back(cluster) 
+
+    interpolated_photons = interpolator.interpolate(clustervector)
+
+    assert interpolated_photons.size == 1
+
+    assert interpolated_photons[0]["x"] == 0
+    assert interpolated_photons[0]["y"] == 0
+    assert interpolated_photons[0]["energy"] == 4
+
+
+
+def test_calculate_eta(): 
+    """Calculate Eta""" 
+    clusters = _aare.ClusterVector_Cluster3x3i() 
+    clusters.push_back(_aare.Cluster3x3i(0,0, np.ones(9, dtype=np.int32)))
+    clusters.push_back(_aare.Cluster3x3i(0,0, np.array([1,1,1,2,2,2,3,3,3])))
+
+    eta2 = _aare.calculate_eta2(clusters)
+
+    assert eta2.shape == (2,2)
+    assert eta2[0,0] == 0.5
+    assert eta2[0,1] == 0.5
+    assert eta2[1,0] == 0.5
+    assert eta2[1,1] == 0.6 #1/5
+
+def test_cluster_finder(): 
+    """Test ClusterFinder""" 
+
+    clusterfinder = _aare.ClusterFinder_Cluster3x3i([100,100])
+
+    #frame = np.random.rand(100,100)
+    frame = np.zeros(shape=[100,100])
+
+    clusterfinder.find_clusters(frame)
+
+    clusters = clusterfinder.steal_clusters(False) #conversion does not work
+
+    assert clusters.size == 0
+
+
+
+
+
+
+
+
+

python/tests/test_ClusterFile.py

@@ -0,0 +1,64 @@
+
+import pytest 
+import numpy as np
+import boost_histogram as bh
+import time
+from pathlib import Path
+import pickle
+
+from aare import ClusterFile
+from conftest import test_data_path
+
+@pytest.mark.files
+def test_cluster_file(test_data_path): 
+    """Test ClusterFile""" 
+    f =  ClusterFile(test_data_path / "clust/single_frame_97_clustrers.clust") 
+    cv = f.read_clusters(10) #conversion does not work
+
+
+    assert cv.frame_number == 135
+    assert cv.size == 10
+
+    #Known data 
+    #frame_number, num_clusters   [135] 97
+    #[  1 200] [0 1 2 3 4 5 6 7 8]
+    #[  2 201] [ 9 10 11 12 13 14 15 16 17]
+    #[  3 202] [18 19 20 21 22 23 24 25 26]
+    #[  4 203] [27 28 29 30 31 32 33 34 35]
+    #[  5 204] [36 37 38 39 40 41 42 43 44]
+    #[  6 205] [45 46 47 48 49 50 51 52 53]
+    #[  7 206] [54 55 56 57 58 59 60 61 62]
+    #[  8 207] [63 64 65 66 67 68 69 70 71]
+    #[  9 208] [72 73 74 75 76 77 78 79 80]
+    #[ 10 209] [81 82 83 84 85 86 87 88 89]
+
+    #conversion to numpy array
+    arr = np.array(cv, copy = False)
+    
+    assert arr.size == 10
+    for i in range(10):
+        assert arr[i]['x'] == i+1
+
+@pytest.mark.files
+def test_read_clusters_and_fill_histogram(test_data_path): 
+    # Create the histogram
+    n_bins = 100
+    xmin = -100
+    xmax = 1e4
+    hist_aare = bh.Histogram(bh.axis.Regular(n_bins, xmin, xmax))
+
+    fname = test_data_path / "clust/beam_En700eV_-40deg_300V_10us_d0_f0_100.clust"
+
+    #Read clusters and fill the histogram with pixel values
+    with ClusterFile(fname, chunk_size = 10000) as f:
+        for clusters in f:
+            arr = np.array(clusters, copy = False)
+            hist_aare.fill(arr['data'].flat)
+
+
+    #Load the histogram from the pickle file
+    with open(fname.with_suffix('.pkl'), 'rb') as f:
+        hist_py = pickle.load(f)
+        
+    #Compare the two histograms
+    assert hist_aare == hist_py

python/tests/test_ClusterVector.py

@@ -0,0 +1,54 @@
+import pytest 
+import numpy as np
+import boost_histogram as bh
+import time
+from pathlib import Path
+import pickle
+
+from aare import ClusterFile
+from aare import _aare
+from conftest import test_data_path
+
+
+def test_create_cluster_vector():
+    cv = _aare.ClusterVector_Cluster3x3i()
+    assert cv.cluster_size_x == 3
+    assert cv.cluster_size_y == 3
+    assert cv.size == 0
+
+
+def test_push_back_on_cluster_vector():
+    cv = _aare.ClusterVector_Cluster2x2i()
+    assert cv.cluster_size_x == 2
+    assert cv.cluster_size_y == 2
+    assert cv.size == 0
+
+    cluster = _aare.Cluster2x2i(19, 22, np.ones(4, dtype=np.int32))
+    cv.push_back(cluster)
+    assert cv.size == 1
+
+    arr = np.array(cv, copy=False)
+    assert arr[0]['x'] == 19
+    assert arr[0]['y'] == 22
+
+
+def test_make_a_hitmap_from_cluster_vector():
+    cv = _aare.ClusterVector_Cluster3x3i()
+
+    # Push back 4 clusters with different positions
+    cv.push_back(_aare.Cluster3x3i(0, 0, np.ones(9, dtype=np.int32)))
+    cv.push_back(_aare.Cluster3x3i(1, 1, np.ones(9, dtype=np.int32)))
+    cv.push_back(_aare.Cluster3x3i(1, 1, np.ones(9, dtype=np.int32)))
+    cv.push_back(_aare.Cluster3x3i(2, 2, np.ones(9, dtype=np.int32)))
+
+    ref = np.zeros((5, 5), dtype=np.int32)
+    ref[0,0] = 1
+    ref[1,1] = 2
+    ref[2,2] = 1
+
+
+    img = _aare.hitmap((5,5), cv)
+    # print(img)
+    # print(ref)
+    assert (img == ref).all()
+    

python/tests/test_RawSubFile.py

@@ -0,0 +1,36 @@
+import pytest
+import numpy as np
+from aare import RawSubFile, DetectorType
+
+
+@pytest.mark.files
+def test_read_a_jungfrau_RawSubFile(test_data_path):
+    with RawSubFile(test_data_path / "raw/jungfrau/jungfrau_single_d0_f1_0.raw", DetectorType.Jungfrau, 512, 1024, 16) as f:
+        assert f.frames_in_file == 3
+
+        headers, frames = f.read()
+
+    assert headers.size == 3
+    assert frames.shape == (3, 512, 1024)
+    
+    # Frame numbers in this file should be 4, 5, 6
+    for i,h in zip(range(4,7,1), headers):
+        assert h["frameNumber"] == i
+
+    # Compare to canned data using numpy
+    data = np.load(test_data_path / "raw/jungfrau/jungfrau_single_0.npy")
+    assert np.all(data[3:6] == frames)
+
+@pytest.mark.files
+def test_iterate_over_a_jungfrau_RawSubFile(test_data_path):
+
+    data = np.load(test_data_path / "raw/jungfrau/jungfrau_single_0.npy")
+
+    with RawSubFile(test_data_path / "raw/jungfrau/jungfrau_single_d0_f0_0.raw", DetectorType.Jungfrau, 512, 1024, 16) as f:
+        i = 0
+        for header, frame in f:
+            assert header["frameNumber"] == i+1
+            assert np.all(frame == data[i])
+            i += 1
+        assert i == 3
+        assert header["frameNumber"] == 3

python/tests/test_jungfrau_dat_files.py

@@ -0,0 +1,92 @@
+import pytest
+import numpy as np
+from aare import JungfrauDataFile
+
+@pytest.mark.files
+def test_jfungfrau_dat_read_number_of_frames(test_data_path):
+    with JungfrauDataFile(test_data_path / "dat/AldoJF500k_000000.dat") as dat_file:
+        assert dat_file.total_frames == 24
+
+    with JungfrauDataFile(test_data_path / "dat/AldoJF250k_000000.dat") as dat_file:
+        assert dat_file.total_frames == 53
+    
+    with JungfrauDataFile(test_data_path / "dat/AldoJF65k_000000.dat") as dat_file:
+        assert dat_file.total_frames == 113
+
+
+@pytest.mark.files
+def test_jfungfrau_dat_read_number_of_file(test_data_path):
+    with JungfrauDataFile(test_data_path / "dat/AldoJF500k_000000.dat") as dat_file:
+        assert dat_file.n_files == 4
+
+    with JungfrauDataFile(test_data_path / "dat/AldoJF250k_000000.dat") as dat_file:
+        assert dat_file.n_files == 7
+    
+    with JungfrauDataFile(test_data_path / "dat/AldoJF65k_000000.dat") as dat_file:
+        assert dat_file.n_files == 7
+
+
+@pytest.mark.files
+def test_read_module(test_data_path):
+    """
+    Read all frames from the series of .dat files. Compare to canned data in npz format. 
+    """
+
+    # Read all frames from the .dat file
+    with JungfrauDataFile(test_data_path / "dat/AldoJF500k_000000.dat") as f:
+        header, data = f.read()
+
+    #Sanity check
+    n_frames = 24
+    assert header.size == n_frames
+    assert data.shape == (n_frames, 512, 1024)
+
+    # Read reference data using numpy
+    with np.load(test_data_path / "dat/AldoJF500k.npz") as f:
+        ref_header = f["headers"]
+        ref_data = f["frames"]
+
+    # Check that the data is the same
+    assert np.all(ref_header == header)
+    assert np.all(ref_data == data)
+
+@pytest.mark.files
+def test_read_half_module(test_data_path):
+
+    # Read all frames from the .dat file
+    with JungfrauDataFile(test_data_path / "dat/AldoJF250k_000000.dat") as f:
+        header, data = f.read()
+
+    n_frames = 53
+    assert header.size == n_frames
+    assert data.shape == (n_frames, 256, 1024)
+
+    # Read reference data using numpy
+    with np.load(test_data_path / "dat/AldoJF250k.npz") as f:
+        ref_header = f["headers"]
+        ref_data = f["frames"]
+
+    # Check that the data is the same
+    assert np.all(ref_header == header)
+    assert np.all(ref_data == data)
+
+
+@pytest.mark.files
+def test_read_single_chip(test_data_path):
+
+    # Read all frames from the .dat file
+    with JungfrauDataFile(test_data_path / "dat/AldoJF65k_000000.dat") as f:
+        header, data = f.read()
+
+    n_frames = 113
+    assert header.size == n_frames
+    assert data.shape == (n_frames, 256, 256)
+
+    # Read reference data using numpy
+    with np.load(test_data_path / "dat/AldoJF65k.npz") as f:
+        ref_header = f["headers"]
+        ref_data = f["frames"]
+
+    # Check that the data is the same
+    assert np.all(ref_header == header)
+    assert np.all(ref_data == data)

src/CalculateEta.test.cpp

@@ -0,0 +1,127 @@
+/************************************************
+ * @file CalculateEta.test.cpp
+ * @short test case to calculate_eta2
+ ***********************************************/
+
+#include "aare/CalculateEta.hpp"
+#include "aare/Cluster.hpp"
+#include "aare/ClusterFile.hpp"
+
+// #include "catch.hpp"
+#include <array>
+#include <catch2/catch_all.hpp>
+#include <catch2/catch_test_macros.hpp>
+
+using namespace aare;
+
+using ClusterTypes =
+    std::variant<Cluster<int, 2, 2>, Cluster<int, 3, 3>, Cluster<int, 5, 5>,
+                 Cluster<int, 4, 2>, Cluster<int, 2, 3>>;
+
+auto get_test_parameters() {
+    return GENERATE(
+        std::make_tuple(ClusterTypes{Cluster<int, 2, 2>{0, 0, {1, 2, 3, 1}}},
+                        Eta2<int>{2. / 3, 3. / 4,
+                                  static_cast<int>(corner::cBottomLeft), 7}),
+        std::make_tuple(
+            ClusterTypes{Cluster<int, 3, 3>{0, 0, {1, 2, 3, 4, 5, 6, 1, 2, 7}}},
+            Eta2<int>{6. / 11, 2. / 7, static_cast<int>(corner::cTopRight),
+                      20}),
+        std::make_tuple(ClusterTypes{Cluster<int, 5, 5>{
+                            0, 0, {1, 6, 7, 6, 5, 4, 3, 2, 1, 2, 8, 9, 8,
+                                   1, 4, 5, 6, 7, 8, 4, 1, 1, 1, 1, 1}}},
+                        Eta2<int>{8. / 17, 7. / 15, 9, 30}),
+        std::make_tuple(
+            ClusterTypes{Cluster<int, 4, 2>{0, 0, {1, 4, 7, 2, 5, 6, 4, 3}}},
+            Eta2<int>{4. / 10, 4. / 11, 1, 21}),
+        std::make_tuple(
+            ClusterTypes{Cluster<int, 2, 3>{0, 0, {1, 3, 2, 3, 4, 2}}},
+            Eta2<int>{3. / 5, 2. / 5, 1, 11}));
+}
+
+TEST_CASE("compute_largest_2x2_subcluster", "[eta_calculation]") {
+    auto [cluster, expected_eta] = get_test_parameters();
+
+    auto [sum, index] = std::visit(
+        [](const auto &clustertype) { return clustertype.max_sum_2x2(); },
+        cluster);
+    CHECK(expected_eta.c == index);
+    CHECK(expected_eta.sum == sum);
+}
+
+TEST_CASE("calculate_eta2", "[eta_calculation]") {
+
+    auto [cluster, expected_eta] = get_test_parameters();
+
+    auto eta = std::visit(
+        [](const auto &clustertype) { return calculate_eta2(clustertype); },
+        cluster);
+
+    CHECK(eta.x == expected_eta.x);
+    CHECK(eta.y == expected_eta.y);
+    CHECK(eta.c == expected_eta.c);
+    CHECK(eta.sum == expected_eta.sum);
+}
+
+// 3x3 cluster layout (rotated to match the cBottomLeft enum):
+//  6, 7, 8
+//  3, 4, 5
+//  0, 1, 2
+
+TEST_CASE("Calculate eta2 for a 3x3 int32 cluster with the largest 2x2 sum in "
+          "the bottom left",
+          "[eta_calculation]") {
+
+    // Create a 3x3 cluster
+    Cluster<int32_t, 3, 3> cl;
+    cl.x = 0;
+    cl.y = 0;
+    cl.data[0] = 30;
+    cl.data[1] = 23;
+    cl.data[2] = 5;
+    cl.data[3] = 20;
+    cl.data[4] = 50;
+    cl.data[5] = 3;
+    cl.data[6] = 8;
+    cl.data[7] = 2;
+    cl.data[8] = 3;
+
+    // 8, 2, 3
+    // 20, 50, 3
+    // 30, 23, 5
+
+    auto eta = calculate_eta2(cl);
+    CHECK(eta.c == static_cast<int>(corner::cBottomLeft));
+    CHECK(eta.x == 50.0 / (20 + 50)); // 4/(3+4)
+    CHECK(eta.y == 50.0 / (23 + 50)); // 4/(1+4)
+    CHECK(eta.sum == 30 + 23 + 20 + 50);
+}
+
+TEST_CASE("Calculate eta2 for a 3x3 int32 cluster with the largest 2x2 sum in "
+          "the top left",
+          "[eta_calculation]") {
+
+    // Create a 3x3 cluster
+    Cluster<int32_t, 3, 3> cl;
+    cl.x = 0;
+    cl.y = 0;
+    cl.data[0] = 8;
+    cl.data[1] = 12;
+    cl.data[2] = 5;
+    cl.data[3] = 77;
+    cl.data[4] = 80;
+    cl.data[5] = 3;
+    cl.data[6] = 82;
+    cl.data[7] = 91;
+    cl.data[8] = 3;
+
+    // 82, 91, 3
+    // 77, 80, 3
+    // 8, 12, 5
+
+    auto eta = calculate_eta2(cl);
+    CHECK(eta.c == static_cast<int>(corner::cTopLeft));
+    CHECK(eta.x == 80. / (77 + 80));  // 4/(3+4)
+    CHECK(eta.y == 91.0 / (91 + 80)); // 7/(7+4)
+    CHECK(eta.sum == 77 + 80 + 82 + 91);
+}

src/Cluster.test.cpp

@@ -0,0 +1,21 @@
+/************************************************
+ * @file test-Cluster.cpp
+ * @short test case for generic Cluster, ClusterVector, and calculate_eta2
+ ***********************************************/
+
+#include "aare/Cluster.hpp"
+#include "aare/CalculateEta.hpp"
+#include "aare/ClusterFile.hpp"
+
+// #include "catch.hpp"
+#include <array>
+#include <catch2/catch_all.hpp>
+#include <catch2/catch_test_macros.hpp>
+
+using namespace aare;
+
+TEST_CASE("Test sum of Cluster", "[.cluster]") {
+    Cluster<int, 2, 2> cluster{0, 0, {1, 2, 3, 4}};
+
+    CHECK(cluster.sum() == 10);
+}

src/ClusterFile.cpp

@@ -1,442 +0,0 @@
-#include "aare/ClusterFile.hpp"
-
-#include <algorithm>
-
-namespace aare {
-
-ClusterFile::ClusterFile(const std::filesystem::path &fname, size_t chunk_size,
-                         const std::string &mode)
-    : m_chunk_size(chunk_size), m_mode(mode) {
-
-    if (mode == "r") {
-        fp = fopen(fname.c_str(), "rb");
-        if (!fp) {
-            throw std::runtime_error("Could not open file for reading: " +
-                                     fname.string());
-        }
-    } else if (mode == "w") {
-        fp = fopen(fname.c_str(), "wb");
-        if (!fp) {
-            throw std::runtime_error("Could not open file for writing: " +
-                                     fname.string());
-        }
-    } else if (mode == "a") {
-        fp = fopen(fname.c_str(), "ab");
-        if (!fp) {
-            throw std::runtime_error("Could not open file for appending: " +
-                                     fname.string());
-        }
-    } else {
-        throw std::runtime_error("Unsupported mode: " + mode);
-    }
-}
-
-ClusterFile::~ClusterFile() { close(); }
-
-void ClusterFile::close() {
-    if (fp) {
-        fclose(fp);
-        fp = nullptr;
-    }
-}
-
-void ClusterFile::write_frame(const ClusterVector<int32_t> &clusters) {
-    if (m_mode != "w" && m_mode != "a") {
-        throw std::runtime_error("File not opened for writing");
-    }
-    if (!(clusters.cluster_size_x() == 3) &&
-        !(clusters.cluster_size_y() == 3)) {
-        throw std::runtime_error("Only 3x3 clusters are supported");
-    }
-    int32_t frame_number = clusters.frame_number();
-    fwrite(&frame_number, sizeof(frame_number), 1, fp);
-    uint32_t n_clusters = clusters.size();
-    fwrite(&n_clusters, sizeof(n_clusters), 1, fp);
-    fwrite(clusters.data(), clusters.item_size(), clusters.size(), fp);
-}
-
-ClusterVector<int32_t> ClusterFile::read_clusters(size_t n_clusters) {
-    if (m_mode != "r") {
-        throw std::runtime_error("File not opened for reading");
-    }
-    
-    ClusterVector<int32_t> clusters(3,3, n_clusters);
-
-    int32_t iframe = 0; // frame number needs to be 4 bytes!
-    size_t nph_read = 0;
-    uint32_t nn = m_num_left;
-    uint32_t nph = m_num_left; // number of clusters in frame needs to be 4
-
-    // auto buf = reinterpret_cast<Cluster3x3 *>(clusters.data());
-    auto buf = clusters.data();
-    // if there are photons left from previous frame read them first
-    if (nph) {
-        if (nph > n_clusters) {
-            // if we have more photons left in the frame then photons to read we
-            // read directly the requested number
-            nn = n_clusters;
-        } else {
-            nn = nph;
-        }
-        nph_read += fread((buf + nph_read*clusters.item_size()),
-                          clusters.item_size(), nn, fp);
-        m_num_left = nph - nn; // write back the number of photons left
-    }
-
-    if (nph_read < n_clusters) {
-        // keep on reading frames and photons until reaching n_clusters
-        while (fread(&iframe, sizeof(iframe), 1, fp)) {
-            // read number of clusters in frame
-            if (fread(&nph, sizeof(nph), 1, fp)) {
-                if (nph > (n_clusters - nph_read))
-                    nn = n_clusters - nph_read;
-                else
-                    nn = nph;
-
-                nph_read += fread((buf + nph_read*clusters.item_size()),
-                                  clusters.item_size(), nn, fp);
-                m_num_left = nph - nn;
-            }
-            if (nph_read >= n_clusters)
-                break;
-        }
-    }
-
-    // Resize the vector to the number of clusters.
-    // No new allocation, only change bounds.
-    clusters.resize(nph_read);
-    return clusters;
-}
-
-ClusterVector<int32_t> ClusterFile::read_frame() {
-    if (m_mode != "r") {
-        throw std::runtime_error("File not opened for reading");
-    }
-    if (m_num_left) {
-        throw std::runtime_error(
-            "There are still photons left in the last frame");
-    }
-    int32_t frame_number;
-    if (fread(&frame_number, sizeof(frame_number), 1, fp) != 1) {
-        throw std::runtime_error("Could not read frame number");
-    }
-
-    int32_t n_clusters; // Saved as 32bit integer in the cluster file
-    if (fread(&n_clusters, sizeof(n_clusters), 1, fp) != 1) {
-        throw std::runtime_error("Could not read number of clusters");
-    }
-    // std::vector<Cluster3x3> clusters(n_clusters);
-    ClusterVector<int32_t> clusters(3, 3, n_clusters);
-    clusters.set_frame_number(frame_number);
-
-    if (fread(clusters.data(), clusters.item_size(), n_clusters, fp) !=
-        static_cast<size_t>(n_clusters)) {
-        throw std::runtime_error("Could not read clusters");
-    }
-    clusters.resize(n_clusters);
-    return clusters;
-}
-
-
-// std::vector<Cluster3x3> ClusterFile::read_cluster_with_cut(size_t n_clusters,
-//                                                            double *noise_map,
-//                                                            int nx, int ny) {
-//     if (m_mode != "r") {
-//         throw std::runtime_error("File not opened for reading");
-//     }
-//     std::vector<Cluster3x3> clusters(n_clusters);
-//     // size_t read_clusters_with_cut(FILE *fp, size_t n_clusters, Cluster *buf,
-//     //                               uint32_t *n_left, double *noise_map, int
-//     //                               nx, int ny) {
-//     int iframe = 0;
-//     //     uint32_t nph = *n_left;
-//     uint32_t nph = m_num_left;
-//     //     uint32_t nn = *n_left;
-//     uint32_t nn = m_num_left;
-//     size_t nph_read = 0;
-
-//     int32_t t2max, tot1;
-//     int32_t tot3;
-//     // Cluster *ptr = buf;
-//     Cluster3x3 *ptr = clusters.data();
-//     int good = 1;
-//     double noise;
-//     // read photons left from previous frame
-//     if (noise_map)
-//         printf("Using noise map\n");
-
-//     if (nph) {
-//         if (nph > n_clusters) {
-//             // if we have more photons left in the frame then photons to
-//             // read we read directly the requested number
-//             nn = n_clusters;
-//         } else {
-//             nn = nph;
-//         }
-//         for (size_t iph = 0; iph < nn; iph++) {
-//             // read photons 1 by 1
-//             size_t n_read =
-//                 fread(reinterpret_cast<void *>(ptr), sizeof(Cluster3x3), 1, fp);
-//             if (n_read != 1) {
-//                 clusters.resize(nph_read);
-//                 return clusters;
-//             }
-//             // TODO! error handling on read
-//             good = 1;
-//             if (noise_map) {
-//                 if (ptr->x >= 0 && ptr->x < nx && ptr->y >= 0 && ptr->y < ny) {
-//                     tot1 = ptr->data[4];
-//                     analyze_cluster(*ptr, &t2max, &tot3, NULL, NULL, NULL, NULL,
-//                                     NULL);
-//                     noise = noise_map[ptr->y * nx + ptr->x];
-//                     if (tot1 > noise || t2max > 2 * noise || tot3 > 3 * noise) {
-//                         ;
-//                     } else {
-//                         good = 0;
-//                         printf("%d %d %f %d %d %d\n", ptr->x, ptr->y, noise,
-//                                tot1, t2max, tot3);
-//                     }
-//                 } else {
-//                     printf("Bad pixel number %d %d\n", ptr->x, ptr->y);
-//                     good = 0;
-//                 }
-//             }
-//             if (good) {
-//                 ptr++;
-//                 nph_read++;
-//             }
-//             (m_num_left)--;
-//             if (nph_read >= n_clusters)
-//                 break;
-//         }
-//     }
-//     if (nph_read < n_clusters) {
-//         //         // keep on reading frames and photons until reaching
-//         //         n_clusters
-//         while (fread(&iframe, sizeof(iframe), 1, fp)) {
-//             //             // printf("%d\n",nph_read);
-
-//             if (fread(&nph, sizeof(nph), 1, fp)) {
-//                 //                 // printf("** %d\n",nph);
-//                 m_num_left = nph;
-//                 for (size_t iph = 0; iph < nph; iph++) {
-//                     //                     // read photons 1 by 1
-//                     size_t n_read = fread(reinterpret_cast<void *>(ptr),
-//                                           sizeof(Cluster3x3), 1, fp);
-//                     if (n_read != 1) {
-//                         clusters.resize(nph_read);
-//                         return clusters;
-//                         // return nph_read;
-//                     }
-//                     good = 1;
-//                     if (noise_map) {
-//                         if (ptr->x >= 0 && ptr->x < nx && ptr->y >= 0 &&
-//                             ptr->y < ny) {
-//                             tot1 = ptr->data[4];
-//                             analyze_cluster(*ptr, &t2max, &tot3, NULL, NULL,
-//                                             NULL, NULL, NULL);
-//                             // noise = noise_map[ptr->y * nx + ptr->x];
-//                             noise = noise_map[ptr->y + ny * ptr->x];
-//                             if (tot1 > noise || t2max > 2 * noise ||
-//                                 tot3 > 3 * noise) {
-//                                 ;
-//                             } else
-//                                 good = 0;
-//                         } else {
-//                             printf("Bad pixel number %d %d\n", ptr->x, ptr->y);
-//                             good = 0;
-//                         }
-//                     }
-//                     if (good) {
-//                         ptr++;
-//                         nph_read++;
-//                     }
-//                     (m_num_left)--;
-//                     if (nph_read >= n_clusters)
-//                         break;
-//                 }
-//             }
-//             if (nph_read >= n_clusters)
-//                 break;
-//         }
-//     }
-//     // printf("%d\n",nph_read);
-//     clusters.resize(nph_read);
-//     return clusters;
-// }
-
-NDArray<double, 2> calculate_eta2(ClusterVector<int> &clusters) {
-    //TOTO! make work with 2x2 clusters
-    NDArray<double, 2> eta2({static_cast<int64_t>(clusters.size()), 2});
-    for (size_t i = 0; i < clusters.size(); i++) {
-        auto e = calculate_eta2(clusters.at<Cluster3x3>(i));
-        eta2(i, 0) = e.x;
-        eta2(i, 1) = e.y;
-    }
-    return eta2;
-}
-
-/** 
- * @brief Calculate the eta2 values for a 3x3 cluster and return them in a Eta2 struct
- * containing etay, etax and the corner of the cluster. 
-*/
-Eta2 calculate_eta2(Cluster3x3 &cl) {
-    Eta2 eta{};
-
-    std::array<int32_t, 4> tot2;
-    tot2[0] = cl.data[0] + cl.data[1] + cl.data[3] + cl.data[4];
-    tot2[1] = cl.data[1] + cl.data[2] + cl.data[4] + cl.data[5];
-    tot2[2] = cl.data[3] + cl.data[4] + cl.data[6] + cl.data[7];
-    tot2[3] = cl.data[4] + cl.data[5] + cl.data[7] + cl.data[8];
-
-    auto c = std::max_element(tot2.begin(), tot2.end()) - tot2.begin();
-
-    switch (c) {
-    case cBottomLeft:
-        if ((cl.data[3] + cl.data[4]) != 0)
-            eta.x =
-                static_cast<double>(cl.data[4]) / (cl.data[3] + cl.data[4]);
-        if ((cl.data[1] + cl.data[4]) != 0)
-            eta.y =
-                static_cast<double>(cl.data[4]) / (cl.data[1] + cl.data[4]);
-        eta.c = cBottomLeft;
-        break;
-    case cBottomRight:
-        if ((cl.data[2] + cl.data[5]) != 0)
-            eta.x =
-                static_cast<double>(cl.data[5]) / (cl.data[4] + cl.data[5]);
-        if ((cl.data[1] + cl.data[4]) != 0)
-            eta.y =
-                static_cast<double>(cl.data[4]) / (cl.data[1] + cl.data[4]);
-        eta.c = cBottomRight;
-        break;
-    case cTopLeft:
-        if ((cl.data[7] + cl.data[4]) != 0)
-            eta.x =
-                static_cast<double>(cl.data[4]) / (cl.data[3] + cl.data[4]);
-        if ((cl.data[7] + cl.data[4]) != 0)
-            eta.y =
-                static_cast<double>(cl.data[7]) / (cl.data[7] + cl.data[4]);
-        eta.c = cTopLeft;
-        break;
-    case cTopRight:
-        if ((cl.data[5] + cl.data[4]) != 0)
-            eta.x =
-                static_cast<double>(cl.data[5]) / (cl.data[5] + cl.data[4]);
-        if ((cl.data[7] + cl.data[4]) != 0)
-            eta.y =
-                static_cast<double>(cl.data[7]) / (cl.data[7] + cl.data[4]);
-        eta.c = cTopRight;
-        break;
-    // no default to allow compiler to warn about missing cases
-    }
-    return eta;
-}
-
-int analyze_cluster(Cluster3x3 &cl, int32_t *t2, int32_t *t3, char *quad,
-                    double *eta2x, double *eta2y, double *eta3x,
-                    double *eta3y) {
-
-    return analyze_data(cl.data, t2, t3, quad, eta2x, eta2y, eta3x, eta3y);
-}
-
-int analyze_data(int32_t *data, int32_t *t2, int32_t *t3, char *quad,
-                 double *eta2x, double *eta2y, double *eta3x, double *eta3y) {
-
-    int ok = 1;
-
-    int32_t tot2[4];
-    int32_t t2max = 0;
-    char c = 0;
-    int32_t val, tot3;
-
-    tot3 = 0;
-    for (int i = 0; i < 4; i++)
-        tot2[i] = 0;
-
-    for (int ix = 0; ix < 3; ix++) {
-        for (int iy = 0; iy < 3; iy++) {
-            val = data[iy * 3 + ix];
-            //	printf ("%d ",data[iy * 3 + ix]);
-            tot3 += val;
-            if (ix <= 1 && iy <= 1)
-                tot2[cBottomLeft] += val;
-            if (ix >= 1 && iy <= 1)
-                tot2[cBottomRight] += val;
-            if (ix <= 1 && iy >= 1)
-                tot2[cTopLeft] += val;
-            if (ix >= 1 && iy >= 1)
-                tot2[cTopRight] += val;
-        }
-        //	printf ("\n");
-    }
-    // printf ("\n");
-
-    if (t2 || quad) {
-
-        t2max = tot2[0];
-        c = cBottomLeft;
-        for (int i = 1; i < 4; i++) {
-            if (tot2[i] > t2max) {
-                t2max = tot2[i];
-                c = i;
-            }
-        }
-        // printf("*** %d %d %d %d --
-        // %d\n",tot2[0],tot2[1],tot2[2],tot2[3],t2max);
-        if (quad)
-            *quad = c;
-        if (t2)
-            *t2 = t2max;
-    }
-
-    if (t3)
-        *t3 = tot3;
-
-    if (eta2x || eta2y) {
-        if (eta2x)
-            *eta2x = 0;
-        if (eta2y)
-            *eta2y = 0;
-        switch (c) {
-        case cBottomLeft:
-            if (eta2x && (data[3] + data[4]) != 0)
-                *eta2x = static_cast<double>(data[4]) / (data[3] + data[4]);
-            if (eta2y && (data[1] + data[4]) != 0)
-                *eta2y = static_cast<double>(data[4]) / (data[1] + data[4]);
-            break;
-        case cBottomRight:
-            if (eta2x && (data[2] + data[5]) != 0)
-                *eta2x = static_cast<double>(data[5]) / (data[4] + data[5]);
-            if (eta2y && (data[1] + data[4]) != 0)
-                *eta2y = static_cast<double>(data[4]) / (data[1] + data[4]);
-            break;
-        case cTopLeft:
-            if (eta2x && (data[7] + data[4]) != 0)
-                *eta2x = static_cast<double>(data[4]) / (data[3] + data[4]);
-            if (eta2y && (data[7] + data[4]) != 0)
-                *eta2y = static_cast<double>(data[7]) / (data[7] + data[4]);
-            break;
-        case cTopRight:
-            if (eta2x && t2max != 0)
-                *eta2x = static_cast<double>(data[5]) / (data[5] + data[4]);
-            if (eta2y && t2max != 0)
-                *eta2y = static_cast<double>(data[7]) / (data[7] + data[4]);
-            break;
-        default:;
-        }
-    }
-
-    if (eta3x || eta3y) {
-        if (eta3x && (data[3] + data[4] + data[5]) != 0)
-            *eta3x = static_cast<double>(-data[3] + data[3 + 2]) /
-                     (data[3] + data[4] + data[5]);
-        if (eta3y && (data[1] + data[4] + data[7]) != 0)
-            *eta3y = static_cast<double>(-data[1] + data[2 * 3 + 1]) /
-                     (data[1] + data[4] + data[7]);
-    }
-
-    return ok;
-}
-
-} // namespace aare

src/ClusterFile.test.cpp

@@ -0,0 +1,351 @@
+#include "aare/ClusterFile.hpp"
+#include "test_config.hpp"
+
+#include "aare/defs.hpp"
+#include <algorithm>
+#include <catch2/catch_test_macros.hpp>
+#include <filesystem>
+
+using aare::Cluster;
+using aare::ClusterFile;
+using aare::ClusterVector;
+
+
+TEST_CASE("Read one frame from a cluster file", "[.files]") {
+    //We know that the frame has 97 clusters
+    auto fpath = test_data_path() / "clust" / "single_frame_97_clustrers.clust";
+    REQUIRE(std::filesystem::exists(fpath));
+
+    ClusterFile<Cluster<int32_t, 3, 3>> f(fpath);
+    auto clusters = f.read_frame();
+    CHECK(clusters.size() == 97);
+    CHECK(clusters.frame_number() == 135);
+    CHECK(clusters[0].x == 1);
+    CHECK(clusters[0].y == 200);
+    int32_t expected_cluster_data[] = {0, 1, 2, 3, 4, 5, 6, 7, 8};
+    CHECK(std::equal(std::begin(clusters[0].data), std::end(clusters[0].data),
+                     std::begin(expected_cluster_data)));
+}
+
+
+TEST_CASE("Read one frame using ROI", "[.files]") {
+    // We know that the frame has 97 clusters
+    auto fpath = test_data_path() / "clust" / "single_frame_97_clustrers.clust";
+    REQUIRE(std::filesystem::exists(fpath));
+
+    ClusterFile<Cluster<int32_t, 3, 3>> f(fpath);
+    aare::ROI roi;
+    roi.xmin = 0;
+    roi.xmax = 50;
+    roi.ymin = 200;
+    roi.ymax = 249;
+    f.set_roi(roi);
+    auto clusters = f.read_frame();
+    REQUIRE(clusters.size() == 49);
+    REQUIRE(clusters.frame_number() == 135);
+
+    // Check that all clusters are within the ROI
+    for (size_t i = 0; i < clusters.size(); i++) {
+        auto c = clusters[i];
+        REQUIRE(c.x >= roi.xmin);
+        REQUIRE(c.x <= roi.xmax);
+        REQUIRE(c.y >= roi.ymin);
+        REQUIRE(c.y <= roi.ymax);
+    }
+
+    CHECK(clusters[0].x == 1);
+    CHECK(clusters[0].y == 200);
+    int32_t expected_cluster_data[] = {0, 1, 2, 3, 4, 5, 6, 7, 8};
+    CHECK(std::equal(std::begin(clusters[0].data), std::end(clusters[0].data),
+                     std::begin(expected_cluster_data)));
+}
+
+
+
+TEST_CASE("Read clusters from single frame file", "[.files]") {
+
+    //  frame_number, num_clusters   [135] 97
+    // [  1 200] [0 1 2 3 4 5 6 7 8]
+    // [  2 201] [ 9 10 11 12 13 14 15 16 17]
+    // [  3 202] [18 19 20 21 22 23 24 25 26]
+    // [  4 203] [27 28 29 30 31 32 33 34 35]
+    // [  5 204] [36 37 38 39 40 41 42 43 44]
+    // [  6 205] [45 46 47 48 49 50 51 52 53]
+    // [  7 206] [54 55 56 57 58 59 60 61 62]
+    // [  8 207] [63 64 65 66 67 68 69 70 71]
+    // [  9 208] [72 73 74 75 76 77 78 79 80]
+    // [ 10 209] [81 82 83 84 85 86 87 88 89]
+    // [ 11 210] [90 91 92 93 94 95 96 97 98]
+    // [ 12 211] [ 99 100 101 102 103 104 105 106 107]
+    // [ 13 212] [108 109 110 111 112 113 114 115 116]
+    // [ 14 213] [117 118 119 120 121 122 123 124 125]
+    // [ 15 214] [126 127 128 129 130 131 132 133 134]
+    // [ 16 215] [135 136 137 138 139 140 141 142 143]
+    // [ 17 216] [144 145 146 147 148 149 150 151 152]
+    // [ 18 217] [153 154 155 156 157 158 159 160 161]
+    // [ 19 218] [162 163 164 165 166 167 168 169 170]
+    // [ 20 219] [171 172 173 174 175 176 177 178 179]
+    // [ 21 220] [180 181 182 183 184 185 186 187 188]
+    // [ 22 221] [189 190 191 192 193 194 195 196 197]
+    // [ 23 222] [198 199 200 201 202 203 204 205 206]
+    // [ 24 223] [207 208 209 210 211 212 213 214 215]
+    // [ 25 224] [216 217 218 219 220 221 222 223 224]
+    // [ 26 225] [225 226 227 228 229 230 231 232 233]
+    // [ 27 226] [234 235 236 237 238 239 240 241 242]
+    // [ 28 227] [243 244 245 246 247 248 249 250 251]
+    // [ 29 228] [252 253 254 255 256 257 258 259 260]
+    // [ 30 229] [261 262 263 264 265 266 267 268 269]
+    // [ 31 230] [270 271 272 273 274 275 276 277 278]
+    // [ 32 231] [279 280 281 282 283 284 285 286 287]
+    // [ 33 232] [288 289 290 291 292 293 294 295 296]
+    // [ 34 233] [297 298 299 300 301 302 303 304 305]
+    // [ 35 234] [306 307 308 309 310 311 312 313 314]
+    // [ 36 235] [315 316 317 318 319 320 321 322 323]
+    // [ 37 236] [324 325 326 327 328 329 330 331 332]
+    // [ 38 237] [333 334 335 336 337 338 339 340 341]
+    // [ 39 238] [342 343 344 345 346 347 348 349 350]
+    // [ 40 239] [351 352 353 354 355 356 357 358 359]
+    // [ 41 240] [360 361 362 363 364 365 366 367 368]
+    // [ 42 241] [369 370 371 372 373 374 375 376 377]
+    // [ 43 242] [378 379 380 381 382 383 384 385 386]
+    // [ 44 243] [387 388 389 390 391 392 393 394 395]
+    // [ 45 244] [396 397 398 399 400 401 402 403 404]
+    // [ 46 245] [405 406 407 408 409 410 411 412 413]
+    // [ 47 246] [414 415 416 417 418 419 420 421 422]
+    // [ 48 247] [423 424 425 426 427 428 429 430 431]
+    // [ 49 248] [432 433 434 435 436 437 438 439 440]
+    // [ 50 249] [441 442 443 444 445 446 447 448 449]
+    // [ 51 250] [450 451 452 453 454 455 456 457 458]
+    // [ 52 251] [459 460 461 462 463 464 465 466 467]
+    // [ 53 252] [468 469 470 471 472 473 474 475 476]
+    // [ 54 253] [477 478 479 480 481 482 483 484 485]
+    // [ 55 254] [486 487 488 489 490 491 492 493 494]
+    // [ 56 255] [495 496 497 498 499 500 501 502 503]
+    // [ 57 256] [504 505 506 507 508 509 510 511 512]
+    // [ 58 257] [513 514 515 516 517 518 519 520 521]
+    // [ 59 258] [522 523 524 525 526 527 528 529 530]
+    // [ 60 259] [531 532 533 534 535 536 537 538 539]
+    // [ 61 260] [540 541 542 543 544 545 546 547 548]
+    // [ 62 261] [549 550 551 552 553 554 555 556 557]
+    // [ 63 262] [558 559 560 561 562 563 564 565 566]
+    // [ 64 263] [567 568 569 570 571 572 573 574 575]
+    // [ 65 264] [576 577 578 579 580 581 582 583 584]
+    // [ 66 265] [585 586 587 588 589 590 591 592 593]
+    // [ 67 266] [594 595 596 597 598 599 600 601 602]
+    // [ 68 267] [603 604 605 606 607 608 609 610 611]
+    // [ 69 268] [612 613 614 615 616 617 618 619 620]
+    // [ 70 269] [621 622 623 624 625 626 627 628 629]
+    // [ 71 270] [630 631 632 633 634 635 636 637 638]
+    // [ 72 271] [639 640 641 642 643 644 645 646 647]
+    // [ 73 272] [648 649 650 651 652 653 654 655 656]
+    // [ 74 273] [657 658 659 660 661 662 663 664 665]
+    // [ 75 274] [666 667 668 669 670 671 672 673 674]
+    // [ 76 275] [675 676 677 678 679 680 681 682 683]
+    // [ 77 276] [684 685 686 687 688 689 690 691 692]
+    // [ 78 277] [693 694 695 696 697 698 699 700 701]
+    // [ 79 278] [702 703 704 705 706 707 708 709 710]
+    // [ 80 279] [711 712 713 714 715 716 717 718 719]
+    // [ 81 280] [720 721 722 723 724 725 726 727 728]
+    // [ 82 281] [729 730 731 732 733 734 735 736 737]
+    // [ 83 282] [738 739 740 741 742 743 744 745 746]
+    // [ 84 283] [747 748 749 750 751 752 753 754 755]
+    // [ 85 284] [756 757 758 759 760 761 762 763 764]
+    // [ 86 285] [765 766 767 768 769 770 771 772 773]
+    // [ 87 286] [774 775 776 777 778 779 780 781 782]
+    // [ 88 287] [783 784 785 786 787 788 789 790 791]
+    // [ 89 288] [792 793 794 795 796 797 798 799 800]
+    // [ 90 289] [801 802 803 804 805 806 807 808 809]
+    // [ 91 290] [810 811 812 813 814 815 816 817 818]
+    // [ 92 291] [819 820 821 822 823 824 825 826 827]
+    // [ 93 292] [828 829 830 831 832 833 834 835 836]
+    // [ 94 293] [837 838 839 840 841 842 843 844 845]
+    // [ 95 294] [846 847 848 849 850 851 852 853 854]
+    // [ 96 295] [855 856 857 858 859 860 861 862 863]
+    // [ 97 296] [864 865 866 867 868 869 870 871 872]
+
+    auto fpath = test_data_path() / "clust" / "single_frame_97_clustrers.clust";
+
+    REQUIRE(std::filesystem::exists(fpath));
+
+    SECTION("Read fewer clusters than available") {
+        ClusterFile<Cluster<int32_t, 3, 3>> f(fpath);
+        auto clusters = f.read_clusters(50);
+        REQUIRE(clusters.size() == 50);
+        REQUIRE(clusters.frame_number() == 135);
+        int32_t expected_cluster_data[] = {0, 1, 2, 3, 4, 5, 6, 7, 8};
+        REQUIRE(clusters[0].x == 1);
+        REQUIRE(clusters[0].y == 200);
+        CHECK(std::equal(std::begin(clusters[0].data),
+                         std::end(clusters[0].data),
+                         std::begin(expected_cluster_data)));
+    }
+    SECTION("Read more clusters than available") {
+        ClusterFile<Cluster<int32_t, 3, 3>> f(fpath);
+        // 100 is the maximum number of clusters read
+        auto clusters = f.read_clusters(100);
+        REQUIRE(clusters.size() == 97);
+        REQUIRE(clusters.frame_number() == 135);
+        int32_t expected_cluster_data[] = {0, 1, 2, 3, 4, 5, 6, 7, 8};
+        REQUIRE(clusters[0].x == 1);
+        REQUIRE(clusters[0].y == 200);
+        CHECK(std::equal(std::begin(clusters[0].data),
+                         std::end(clusters[0].data),
+                         std::begin(expected_cluster_data)));
+    }
+    SECTION("Read all clusters") {
+        ClusterFile<Cluster<int32_t, 3, 3>> f(fpath);
+        auto clusters = f.read_clusters(97);
+        REQUIRE(clusters.size() == 97);
+        REQUIRE(clusters.frame_number() == 135);
+        REQUIRE(clusters[0].x == 1);
+        REQUIRE(clusters[0].y == 200);
+        int32_t expected_cluster_data[] = {0, 1, 2, 3, 4, 5, 6, 7, 8};
+        CHECK(std::equal(std::begin(clusters[0].data),
+                         std::end(clusters[0].data),
+                         std::begin(expected_cluster_data)));
+    }
+}
+
+TEST_CASE("Read clusters from single frame file with ROI", "[.files]") {
+    auto fpath = test_data_path() / "clust" / "single_frame_97_clustrers.clust";
+    REQUIRE(std::filesystem::exists(fpath));
+
+    ClusterFile<Cluster<int32_t, 3, 3>> f(fpath);
+
+    aare::ROI roi;
+    roi.xmin = 0;
+    roi.xmax = 50;
+    roi.ymin = 200;
+    roi.ymax = 249;
+    f.set_roi(roi);
+
+    auto clusters = f.read_clusters(10);
+
+    CHECK(clusters.size() == 10);
+    CHECK(clusters.frame_number() == 135);
+    CHECK(clusters[0].x == 1);
+    CHECK(clusters[0].y == 200);
+    int32_t expected_cluster_data[] = {0, 1, 2, 3, 4, 5, 6, 7, 8};
+    CHECK(std::equal(std::begin(clusters[0].data), std::end(clusters[0].data),
+                     std::begin(expected_cluster_data)));
+}
+
+TEST_CASE("Read cluster from multiple frame file", "[.files]") {
+
+    using ClusterType = Cluster<double, 2, 2>;
+
+    auto fpath =
+        test_data_path() / "clust" / "Two_frames_2x2double_test_clusters.clust";
+
+    REQUIRE(std::filesystem::exists(fpath));
+
+    // Two_frames_2x2double_test_clusters.clust
+    //  frame number, num_clusters 0, 4
+    //[10, 20], {0. ,0., 0., 0.}
+    //[11, 30], {1., 1., 1., 1.}
+    //[12, 40], {2., 2., 2., 2.}
+    //[13, 50], {3., 3., 3., 3.}
+    //  1,4
+    //[10, 20], {4., 4., 4., 4.}
+    //[11, 30], {5., 5., 5., 5.}
+    //[12, 40], {6., 6., 6., 6.}
+    //[13, 50], {7., 7., 7., 7.}
+
+    SECTION("Read clusters from both frames") {
+        ClusterFile<ClusterType> f(fpath);
+        auto clusters = f.read_clusters(2);
+        REQUIRE(clusters.size() == 2);
+        REQUIRE(clusters.frame_number() == 0);
+
+        auto clusters1 = f.read_clusters(3);
+
+        REQUIRE(clusters1.size() == 3);
+        REQUIRE(clusters1.frame_number() == 1);
+    }
+
+    SECTION("Read all clusters") {
+        ClusterFile<ClusterType> f(fpath);
+        auto clusters = f.read_clusters(8);
+        REQUIRE(clusters.size() == 8);
+        REQUIRE(clusters.frame_number() == 1);
+    }
+
+    SECTION("Read clusters from one frame") {
+        ClusterFile<ClusterType> f(fpath);
+        auto clusters = f.read_clusters(2);
+        REQUIRE(clusters.size() == 2);
+        REQUIRE(clusters.frame_number() == 0);
+
+        auto clusters1 = f.read_clusters(1);
+
+        REQUIRE(clusters1.size() == 1);
+        REQUIRE(clusters1.frame_number() == 0);
+    }
+}
+
+TEST_CASE("Write cluster with potential padding", "[.files][.ClusterFile]") {
+
+    using ClusterType = Cluster<double, 3, 3>;
+
+    REQUIRE(std::filesystem::exists(test_data_path() / "clust"));
+
+    auto fpath = test_data_path() / "clust" / "single_frame_2_clusters.clust";
+
+    ClusterFile<ClusterType> file(fpath, 1000, "w");
+
+    ClusterVector<ClusterType> clustervec(2);
+    int16_t coordinate = 5;
+    clustervec.push_back(ClusterType{
+        coordinate, coordinate, {0., 0., 0., 0., 0., 0., 0., 0., 0.}});
+    clustervec.push_back(ClusterType{
+        coordinate, coordinate, {0., 0., 0., 0., 0., 0., 0., 0., 0.}});
+
+    file.write_frame(clustervec);
+
+    file.close();
+
+    file.open("r");
+
+    auto read_cluster_vector = file.read_frame();
+
+    CHECK(read_cluster_vector.size() == 2);
+    CHECK(read_cluster_vector.frame_number() == 0);
+
+    CHECK(read_cluster_vector[0].x == clustervec[0].x);
+    CHECK(read_cluster_vector[0].y == clustervec[0].y);
+    CHECK(std::equal(
+        clustervec[0].data.begin(), clustervec[0].data.end(),
+        read_cluster_vector[0].data.begin(), [](double a, double b) {
+            return std::abs(a - b) < std::numeric_limits<double>::epsilon();
+        }));
+
+    CHECK(read_cluster_vector[1].x == clustervec[1].x);
+    CHECK(read_cluster_vector[1].y == clustervec[1].y);
+    CHECK(std::equal(
+        clustervec[1].data.begin(), clustervec[1].data.end(),
+        read_cluster_vector[1].data.begin(), [](double a, double b) {
+            return std::abs(a - b) < std::numeric_limits<double>::epsilon();
+        }));
+}
+
+TEST_CASE("Read frame and modify cluster data", "[.files][.ClusterFile]") {
+    auto fpath = test_data_path() / "clust" / "single_frame_97_clustrers.clust";
+    REQUIRE(std::filesystem::exists(fpath));
+
+    ClusterFile<Cluster<int32_t, 3, 3>> f(fpath);
+
+    auto clusters = f.read_frame();
+    CHECK(clusters.size() == 97);
+    CHECK(clusters.frame_number() == 135);
+
+    int32_t expected_cluster_data[] = {0, 1, 2, 3, 4, 5, 6, 7, 8};
+    clusters.push_back(
+        Cluster<int32_t, 3, 3>{0, 0, {0, 1, 2, 3, 4, 5, 6, 7, 8}});
+
+    CHECK(clusters.size() == 98);
+    CHECK(clusters[0].x == 1);
+    CHECK(clusters[0].y == 200);
+
+    CHECK(std::equal(std::begin(clusters[0].data), std::end(clusters[0].data),
+                     std::begin(expected_cluster_data)));
+}

src/ClusterFinder.test.cpp

@@ -1,19 +1,18 @@
 #include "aare/ClusterFinder.hpp"
 #include "aare/Pedestal.hpp"
-#include <catch2/matchers/catch_matchers_floating_point.hpp>
 #include <catch2/catch_test_macros.hpp>
+#include <catch2/matchers/catch_matchers_floating_point.hpp>
 #include <chrono>
 #include <random>
 
 using namespace aare;
 
-//TODO! Find a way to test the cluster finder
-
-
+// TODO! Find a way to test the cluster finder
 
 // class ClusterFinderUnitTest : public ClusterFinder {
 //   public:
-//     ClusterFinderUnitTest(int cluster_sizeX, int cluster_sizeY, double nSigma = 5.0, double threshold = 0.0)
+//     ClusterFinderUnitTest(int cluster_sizeX, int cluster_sizeY, double nSigma
+//     = 5.0, double threshold = 0.0)
 //         : ClusterFinder(cluster_sizeX, cluster_sizeY, nSigma, threshold) {}
 //     double get_c2() { return c2; }
 //     double get_c3() { return c3; }
@@ -37,8 +36,8 @@ using namespace aare;
 //     REQUIRE_THAT(cf.get_c3(), Catch::Matchers::WithinRel(c3, 1e-9));
 // }
 
-TEST_CASE("Construct a cluster finder"){
-    ClusterFinder clusterFinder({400,400}, {3,3});
+TEST_CASE("Construct a cluster finder") {
+    ClusterFinder clusterFinder({400, 400});
     // REQUIRE(clusterFinder.get_cluster_sizeX() == 3);
     // REQUIRE(clusterFinder.get_cluster_sizeY() == 3);
     // REQUIRE(clusterFinder.get_threshold() == 1);
@@ -49,16 +48,17 @@ TEST_CASE("Construct a cluster finder"){
 //     aare::Pedestal pedestal(10, 10, 5);
 //     NDArray<double, 2> frame({10, 10});
 //     frame = 0;
-//     ClusterFinder clusterFinder(3, 3, 1, 1); // 3x3 cluster, 1 nSigma, 1 threshold
+//     ClusterFinder clusterFinder(3, 3, 1, 1); // 3x3 cluster, 1 nSigma, 1
+//     threshold
 
-//     auto clusters = clusterFinder.find_clusters_without_threshold(frame.span(), pedestal);
+//     auto clusters =
+//     clusterFinder.find_clusters_without_threshold(frame.span(), pedestal);
 
 //     REQUIRE(clusters.size() == 0);
 
 //     frame(5, 5) = 10;
-//     clusters = clusterFinder.find_clusters_without_threshold(frame.span(), pedestal);
-//     REQUIRE(clusters.size() == 1);
-//     REQUIRE(clusters[0].x == 5);
+//     clusters = clusterFinder.find_clusters_without_threshold(frame.span(),
+//     pedestal); REQUIRE(clusters.size() == 1); REQUIRE(clusters[0].x == 5);
 //     REQUIRE(clusters[0].y == 5);
 //     for (int i = 0; i < 3; i++) {
 //         for (int j = 0; j < 3; j++) {

src/ClusterFinderMT.test.cpp

@@ -0,0 +1,99 @@
+
+#include "aare/ClusterFinderMT.hpp"
+#include "aare/Cluster.hpp"
+#include "aare/ClusterCollector.hpp"
+#include "aare/File.hpp"
+
+#include "test_config.hpp"
+
+#include <catch2/catch_test_macros.hpp>
+#include <filesystem>
+#include <memory>
+
+using namespace aare;
+
+// wrapper function to access private member variables for testing
+template <typename ClusterType, typename FRAME_TYPE = uint16_t,
+          typename PEDESTAL_TYPE = double>
+class ClusterFinderMTWrapper
+    : public ClusterFinderMT<ClusterType, FRAME_TYPE, PEDESTAL_TYPE> {
+
+  public:
+    ClusterFinderMTWrapper(Shape<2> image_size, PEDESTAL_TYPE nSigma = 5.0,
+                           size_t capacity = 2000, size_t n_threads = 3)
+        : ClusterFinderMT<ClusterType, FRAME_TYPE, PEDESTAL_TYPE>(
+              image_size, nSigma, capacity, n_threads) {}
+
+    size_t get_m_input_queues_size() const {
+        return this->m_input_queues.size();
+    }
+
+    size_t get_m_output_queues_size() const {
+        return this->m_output_queues.size();
+    }
+
+    size_t get_m_cluster_finders_size() const {
+        return this->m_cluster_finders.size();
+    }
+
+    bool m_output_queues_are_empty() const {
+        for (auto &queue : this->m_output_queues) {
+            if (!queue->isEmpty())
+                return false;
+        }
+        return true;
+    }
+
+    bool m_input_queues_are_empty() const {
+        for (auto &queue : this->m_input_queues) {
+            if (!queue->isEmpty())
+                return false;
+        }
+        return true;
+    }
+
+    bool m_sink_is_empty() const { return this->m_sink.isEmpty(); }
+
+    size_t m_sink_size() const { return this->m_sink.sizeGuess(); }
+};
+
+TEST_CASE("multithreaded cluster finder", "[.files][.ClusterFinder]") {
+    auto fpath = "/mnt/sls_det_storage/matterhorn_data/aare_test_data/"
+                 "Moench03new/cu_half_speed_master_4.json";
+
+    File file(fpath);
+
+    size_t n_threads = 2;
+    size_t n_frames_pd = 10;
+
+    using ClusterType = Cluster<int32_t, 3, 3>;
+
+    ClusterFinderMTWrapper<ClusterType> cf(
+        {static_cast<int64_t>(file.rows()), static_cast<int64_t>(file.cols())},
+        5, 2000, n_threads); // no idea what frame type is!!! default uint16_t
+
+    CHECK(cf.get_m_input_queues_size() == n_threads);
+    CHECK(cf.get_m_output_queues_size() == n_threads);
+    CHECK(cf.get_m_cluster_finders_size() == n_threads);
+    CHECK(cf.m_output_queues_are_empty() == true);
+    CHECK(cf.m_input_queues_are_empty() == true);
+
+    for (size_t i = 0; i < n_frames_pd; ++i) {
+        cf.find_clusters(file.read_frame().view<uint16_t>());
+    }
+
+    cf.stop();
+
+    CHECK(cf.m_output_queues_are_empty() == true);
+    CHECK(cf.m_input_queues_are_empty() == true);
+
+    CHECK(cf.m_sink_size() == n_frames_pd);
+    ClusterCollector<ClusterType> clustercollector(&cf);
+
+    clustercollector.stop();
+
+    CHECK(cf.m_sink_size() == 0);
+
+    auto clustervec = clustercollector.steal_clusters();
+    // CHECK(clustervec.size() == ) //dont know how many clusters to expect
+}

src/ClusterVector.test.cpp

@@ -1,21 +1,52 @@
-#include <cstdint>
 #include "aare/ClusterVector.hpp"
+#include <cstdint>
 
-#include <catch2/matchers/catch_matchers_floating_point.hpp>
+#include <catch2/catch_all.hpp>
 #include <catch2/catch_test_macros.hpp>
+#include <catch2/matchers/catch_matchers_floating_point.hpp>
 
+using aare::Cluster;
 using aare::ClusterVector;
 
-struct Cluster_i2x2 {
-    int16_t x;
-    int16_t y;
-    int32_t data[4];
-};
+TEST_CASE("item_size return the size of the cluster stored") {
+    using C1 = Cluster<int32_t, 2, 2>;
+    ClusterVector<C1> cv(4);
+    CHECK(cv.item_size() == sizeof(C1));
 
-TEST_CASE("ClusterVector 2x2 int32_t capacity 4, push back then read") {
-    
+    // Sanity check
+    // 2*2*4 = 16 bytes of data for the cluster
+    //  2*2 = 4 bytes for the x and y coordinates
+    REQUIRE(cv.item_size() == 20);
 
-    ClusterVector<int32_t> cv(2, 2, 4);
+    using C2 = Cluster<int32_t, 3, 3>;
+    ClusterVector<C2> cv2(4);
+    CHECK(cv2.item_size() == sizeof(C2));
+
+    using C3 = Cluster<double, 2, 3>;
+    ClusterVector<C3> cv3(4);
+    CHECK(cv3.item_size() == sizeof(C3));
+
+    using C4 = Cluster<char, 10, 5>;
+    ClusterVector<C4> cv4(4);
+    CHECK(cv4.item_size() == sizeof(C4));
+
+    using C5 = Cluster<int32_t, 2, 3>;
+    ClusterVector<C5> cv5(4);
+    CHECK(cv5.item_size() == sizeof(C5));
+
+    using C6 = Cluster<double, 5, 5>;
+    ClusterVector<C6> cv6(4);
+    CHECK(cv6.item_size() == sizeof(C6));
+
+    using C7 = Cluster<double, 3, 3>;
+    ClusterVector<C7> cv7(4);
+    CHECK(cv7.item_size() == sizeof(C7));
+}
+
+TEST_CASE("ClusterVector 2x2 int32_t capacity 4, push back then read",
+          "[.ClusterVector]") {
+
+    ClusterVector<Cluster<int32_t, 2, 2>> cv(4);
     REQUIRE(cv.capacity() == 4);
     REQUIRE(cv.size() == 0);
     REQUIRE(cv.cluster_size_x() == 2);
@@ -23,112 +54,102 @@ TEST_CASE("ClusterVector 2x2 int32_t capacity 4, push back then read") {
     // int16_t, int16_t, 2x2 int32_t = 20 bytes
     REQUIRE(cv.item_size() == 20);
 
-    //Create a cluster and push back into the vector
-    Cluster_i2x2 c1 = {1, 2, {3, 4, 5, 6}};
-    cv.push_back(c1.x, c1.y, reinterpret_cast<std::byte*>(&c1.data[0]));
+    // Create a cluster and push back into the vector
+    Cluster<int32_t, 2, 2> c1 = {1, 2, {3, 4, 5, 6}};
+    cv.push_back(c1);
     REQUIRE(cv.size() == 1);
     REQUIRE(cv.capacity() == 4);
 
-    //Read the cluster back out using copy. TODO! Can we improve the API?
-    Cluster_i2x2 c2;
-    std::byte *ptr = cv.element_ptr(0);
-    std::copy(ptr, ptr + cv.item_size(), reinterpret_cast<std::byte*>(&c2));
+    auto c2 = cv[0];
 
-    //Check that the data is the same
+    // Check that the data is the same
     REQUIRE(c1.x == c2.x);
     REQUIRE(c1.y == c2.y);
-    for(size_t i = 0; i < 4; i++) {
+    for (size_t i = 0; i < 4; i++) {
         REQUIRE(c1.data[i] == c2.data[i]);
     }
 }
 
-TEST_CASE("Summing 3x1 clusters of int64"){
-    struct Cluster_l3x1{
-        int16_t x;
-        int16_t y;
-        int32_t data[3];
-    };
-
-    ClusterVector<int32_t> cv(3, 1, 2);
+TEST_CASE("Summing 3x1 clusters of int64", "[.ClusterVector]") {
+    ClusterVector<Cluster<int32_t, 3, 1>> cv(2);
     REQUIRE(cv.capacity() == 2);
     REQUIRE(cv.size() == 0);
     REQUIRE(cv.cluster_size_x() == 3);
     REQUIRE(cv.cluster_size_y() == 1);
 
-    //Create a cluster and push back into the vector
-    Cluster_l3x1 c1 = {1, 2, {3, 4, 5}};
-    cv.push_back(c1.x, c1.y, reinterpret_cast<std::byte*>(&c1.data[0]));
+    // Create a cluster and push back into the vector
+    Cluster<int32_t, 3, 1> c1 = {1, 2, {3, 4, 5}};
+    cv.push_back(c1);
     REQUIRE(cv.capacity() == 2);
     REQUIRE(cv.size() == 1);
 
-    Cluster_l3x1 c2 = {6, 7, {8, 9, 10}};
-    cv.push_back(c2.x, c2.y, reinterpret_cast<std::byte*>(&c2.data[0]));
+    Cluster<int32_t, 3, 1> c2 = {6, 7, {8, 9, 10}};
+    cv.push_back(c2);
     REQUIRE(cv.capacity() == 2);
     REQUIRE(cv.size() == 2);
 
-    Cluster_l3x1 c3 = {11, 12, {13, 14, 15}};
-    cv.push_back(c3.x, c3.y, reinterpret_cast<std::byte*>(&c3.data[0]));
+    Cluster<int32_t, 3, 1> c3 = {11, 12, {13, 14, 15}};
+    cv.push_back(c3);
     REQUIRE(cv.capacity() == 4);
     REQUIRE(cv.size() == 3);
 
+    /*
     auto sums = cv.sum();
     REQUIRE(sums.size() == 3);
     REQUIRE(sums[0] == 12);
     REQUIRE(sums[1] == 27);
     REQUIRE(sums[2] == 42);
+    */
 }
 
-TEST_CASE("Storing floats"){
-    struct Cluster_f4x2{
-        int16_t x;
-        int16_t y;
-        float data[8];
-    };
-
-    ClusterVector<float> cv(2, 4, 10);
+TEST_CASE("Storing floats", "[.ClusterVector]") {
+    ClusterVector<Cluster<float, 2, 4>> cv(10);
     REQUIRE(cv.capacity() == 10);
     REQUIRE(cv.size() == 0);
     REQUIRE(cv.cluster_size_x() == 2);
     REQUIRE(cv.cluster_size_y() == 4);
 
-    //Create a cluster and push back into the vector
-    Cluster_f4x2 c1 = {1, 2, {3.0, 4.0, 5.0, 6.0,3.0, 4.0, 5.0, 6.0}};
-    cv.push_back(c1.x, c1.y, reinterpret_cast<std::byte*>(&c1.data[0]));
+    // Create a cluster and push back into the vector
+    Cluster<float, 2, 4> c1 = {1, 2, {3.0, 4.0, 5.0, 6.0, 3.0, 4.0, 5.0, 6.0}};
+    cv.push_back(c1);
     REQUIRE(cv.capacity() == 10);
     REQUIRE(cv.size() == 1);
-    
 
-    Cluster_f4x2 c2 = {6, 7, {8.0, 9.0, 10.0, 11.0,8.0, 9.0, 10.0, 11.0}};
-    cv.push_back(c2.x, c2.y, reinterpret_cast<std::byte*>(&c2.data[0]));
+    Cluster<float, 2, 4> c2 = {
+        6, 7, {8.0, 9.0, 10.0, 11.0, 8.0, 9.0, 10.0, 11.0}};
+    cv.push_back(c2);
     REQUIRE(cv.capacity() == 10);
     REQUIRE(cv.size() == 2);
 
+    /*
     auto sums = cv.sum();
     REQUIRE(sums.size() == 2);
     REQUIRE_THAT(sums[0], Catch::Matchers::WithinAbs(36.0, 1e-6));
     REQUIRE_THAT(sums[1], Catch::Matchers::WithinAbs(76.0, 1e-6));
+    */
 }
 
-TEST_CASE("Push back more than initial capacity"){
-    
-    ClusterVector<int32_t> cv(2, 2, 2);
+TEST_CASE("Push back more than initial capacity", "[.ClusterVector]") {
+
+    ClusterVector<Cluster<int32_t, 2, 2>> cv(2);
     auto initial_data = cv.data();
-    Cluster_i2x2 c1 = {1, 2, {3, 4, 5, 6}};
-    cv.push_back(c1.x, c1.y, reinterpret_cast<std::byte*>(&c1.data[0]));
+    Cluster<int32_t, 2, 2> c1 = {1, 2, {3, 4, 5, 6}};
+    cv.push_back(c1);
     REQUIRE(cv.size() == 1);
     REQUIRE(cv.capacity() == 2);
 
-    Cluster_i2x2 c2 = {6, 7, {8, 9, 10, 11}};
-    cv.push_back(c2.x, c2.y, reinterpret_cast<std::byte*>(&c2.data[0]));
+    Cluster<int32_t, 2, 2> c2 = {6, 7, {8, 9, 10, 11}};
+    cv.push_back(c2);
     REQUIRE(cv.size() == 2);
     REQUIRE(cv.capacity() == 2);
 
-    Cluster_i2x2 c3 = {11, 12, {13, 14, 15, 16}};
-    cv.push_back(c3.x, c3.y, reinterpret_cast<std::byte*>(&c3.data[0]));
-    REQUIRE(cv.size() == 3); 
+    Cluster<int32_t, 2, 2> c3 = {11, 12, {13, 14, 15, 16}};
+    cv.push_back(c3);
+    REQUIRE(cv.size() == 3);
     REQUIRE(cv.capacity() == 4);
 
-    Cluster_i2x2* ptr = reinterpret_cast<Cluster_i2x2*>(cv.data());
+    Cluster<int32_t, 2, 2> *ptr =
+        reinterpret_cast<Cluster<int32_t, 2, 2> *>(cv.data());
     REQUIRE(ptr[0].x == 1);
     REQUIRE(ptr[0].y == 2);
     REQUIRE(ptr[1].x == 6);
@@ -136,29 +157,31 @@ TEST_CASE("Push back more than initial capacity"){
     REQUIRE(ptr[2].x == 11);
     REQUIRE(ptr[2].y == 12);
 
-    //We should have allocated a new buffer, since we outgrew the initial capacity
+    // We should have allocated a new buffer, since we outgrew the initial
+    // capacity
     REQUIRE(initial_data != cv.data());
-
 }
 
-TEST_CASE("Concatenate two cluster vectors where the first has enough capacity"){
-    ClusterVector<int32_t> cv1(2, 2, 12);
-    Cluster_i2x2 c1 = {1, 2, {3, 4, 5, 6}};
-    cv1.push_back(c1.x, c1.y, reinterpret_cast<std::byte*>(&c1.data[0]));
-    Cluster_i2x2 c2 = {6, 7, {8, 9, 10, 11}};
-    cv1.push_back(c2.x, c2.y, reinterpret_cast<std::byte*>(&c2.data[0]));
+TEST_CASE("Concatenate two cluster vectors where the first has enough capacity",
+          "[.ClusterVector]") {
+    ClusterVector<Cluster<int32_t, 2, 2>> cv1(12);
+    Cluster<int32_t, 2, 2> c1 = {1, 2, {3, 4, 5, 6}};
+    cv1.push_back(c1);
+    Cluster<int32_t, 2, 2> c2 = {6, 7, {8, 9, 10, 11}};
+    cv1.push_back(c2);
 
-    ClusterVector<int32_t> cv2(2, 2, 2);
-    Cluster_i2x2 c3 = {11, 12, {13, 14, 15, 16}};
-    cv2.push_back(c3.x, c3.y, reinterpret_cast<std::byte*>(&c3.data[0]));
-    Cluster_i2x2 c4 = {16, 17, {18, 19, 20, 21}};
-    cv2.push_back(c4.x, c4.y, reinterpret_cast<std::byte*>(&c4.data[0]));
+    ClusterVector<Cluster<int32_t, 2, 2>> cv2(2);
+    Cluster<int32_t, 2, 2> c3 = {11, 12, {13, 14, 15, 16}};
+    cv2.push_back(c3);
+    Cluster<int32_t, 2, 2> c4 = {16, 17, {18, 19, 20, 21}};
+    cv2.push_back(c4);
 
     cv1 += cv2;
     REQUIRE(cv1.size() == 4);
     REQUIRE(cv1.capacity() == 12);
 
-    Cluster_i2x2* ptr = reinterpret_cast<Cluster_i2x2*>(cv1.data());
+    Cluster<int32_t, 2, 2> *ptr =
+        reinterpret_cast<Cluster<int32_t, 2, 2> *>(cv1.data());
     REQUIRE(ptr[0].x == 1);
     REQUIRE(ptr[0].y == 2);
     REQUIRE(ptr[1].x == 6);
@@ -169,24 +192,26 @@ TEST_CASE("Concatenate two cluster vectors where the first has enough capacity")
     REQUIRE(ptr[3].y == 17);
 }
 
-TEST_CASE("Concatenate two cluster vectors where we need to allocate"){
-    ClusterVector<int32_t> cv1(2, 2, 2);
-    Cluster_i2x2 c1 = {1, 2, {3, 4, 5, 6}};
-    cv1.push_back(c1.x, c1.y, reinterpret_cast<std::byte*>(&c1.data[0]));
-    Cluster_i2x2 c2 = {6, 7, {8, 9, 10, 11}};
-    cv1.push_back(c2.x, c2.y, reinterpret_cast<std::byte*>(&c2.data[0]));
+TEST_CASE("Concatenate two cluster vectors where we need to allocate",
+          "[.ClusterVector]") {
+    ClusterVector<Cluster<int32_t, 2, 2>> cv1(2);
+    Cluster<int32_t, 2, 2> c1 = {1, 2, {3, 4, 5, 6}};
+    cv1.push_back(c1);
+    Cluster<int32_t, 2, 2> c2 = {6, 7, {8, 9, 10, 11}};
+    cv1.push_back(c2);
 
-    ClusterVector<int32_t> cv2(2, 2, 2);
-    Cluster_i2x2 c3 = {11, 12, {13, 14, 15, 16}};
-    cv2.push_back(c3.x, c3.y, reinterpret_cast<std::byte*>(&c3.data[0]));
-    Cluster_i2x2 c4 = {16, 17, {18, 19, 20, 21}};
-    cv2.push_back(c4.x, c4.y, reinterpret_cast<std::byte*>(&c4.data[0]));
+    ClusterVector<Cluster<int32_t, 2, 2>> cv2(2);
+    Cluster<int32_t, 2, 2> c3 = {11, 12, {13, 14, 15, 16}};
+    cv2.push_back(c3);
+    Cluster<int32_t, 2, 2> c4 = {16, 17, {18, 19, 20, 21}};
+    cv2.push_back(c4);
 
     cv1 += cv2;
     REQUIRE(cv1.size() == 4);
     REQUIRE(cv1.capacity() == 4);
 
-    Cluster_i2x2* ptr = reinterpret_cast<Cluster_i2x2*>(cv1.data());
+    Cluster<int32_t, 2, 2> *ptr =
+        reinterpret_cast<Cluster<int32_t, 2, 2> *>(cv1.data());
     REQUIRE(ptr[0].x == 1);
     REQUIRE(ptr[0].y == 2);
     REQUIRE(ptr[1].x == 6);
@@ -195,4 +220,49 @@ TEST_CASE("Concatenate two cluster vectors where we need to allocate"){
     REQUIRE(ptr[2].y == 12);
     REQUIRE(ptr[3].x == 16);
     REQUIRE(ptr[3].y == 17);
+}
+
+struct ClusterTestData {
+    uint8_t ClusterSizeX;
+    uint8_t ClusterSizeY;
+    std::vector<int64_t> index_map_x;
+    std::vector<int64_t> index_map_y;
+};
+
+TEST_CASE("Gain Map Calculation Index Map", "[.ClusterVector][.gain_map]") {
+
+    auto clustertestdata = GENERATE(
+        ClusterTestData{3,
+                        3,
+                        {-1, 0, 1, -1, 0, 1, -1, 0, 1},
+                        {-1, -1, -1, 0, 0, 0, 1, 1, 1}},
+        ClusterTestData{
+            4,
+            4,
+            {-2, -1, 0, 1, -2, -1, 0, 1, -2, -1, 0, 1, -2, -1, 0, 1},
+            {-2, -2, -2, -2, -1, -1, -1, -1, 0, 0, 0, 0, 1, 1, 1, 1}},
+        ClusterTestData{2, 2, {-1, 0, -1, 0}, {-1, -1, 0, 0}},
+        ClusterTestData{5,
+                        5,
+                        {-2, -1, 0,  1,  2, -2, -1, 0,  1,  2, -2, -1, 0,
+                         1,  2,  -2, -1, 0, 1,  2,  -2, -1, 0, 1,  2},
+                        {-2, -2, -2, -2, -2, -1, -1, -1, -1, -1, 0, 0, 0,
+                         0,  0,  1,  1,  1,  1,  1,  2,  2,  2,  2, 2}});
+
+    uint8_t ClusterSizeX = clustertestdata.ClusterSizeX;
+    uint8_t ClusterSizeY = clustertestdata.ClusterSizeY;
+
+    std::vector<int64_t> index_map_x(ClusterSizeX * ClusterSizeY);
+    std::vector<int64_t> index_map_y(ClusterSizeX * ClusterSizeY);
+
+    int64_t index_cluster_center_x = ClusterSizeX / 2;
+    int64_t index_cluster_center_y = ClusterSizeY / 2;
+
+    for (size_t j = 0; j < ClusterSizeX * ClusterSizeY; j++) {
+        index_map_x[j] = j % ClusterSizeX - index_cluster_center_x;
+        index_map_y[j] = j / ClusterSizeX - index_cluster_center_y;
+    }
+
+    CHECK(index_map_x == clustertestdata.index_map_x);
+    CHECK(index_map_y == clustertestdata.index_map_y);
 }

src/Dtype.cpp

@@ -70,7 +70,7 @@ uint8_t Dtype::bitdepth() const {
 /**
  * @brief Get the number of bytes of the data type
  */
-size_t Dtype::bytes() const { return bitdepth() / 8; }
+size_t Dtype::bytes() const { return bitdepth() / bits_per_byte; }
 
 /**
  * @brief Construct a DType object from a TypeIndex

src/File.cpp

@@ -1,4 +1,5 @@
 #include "aare/File.hpp"
+#include "aare/JungfrauDataFile.hpp"
 #include "aare/NumpyFile.hpp"
 #include "aare/RawFile.hpp"
 
@@ -27,6 +28,8 @@ File::File(const std::filesystem::path &fname, const std::string &mode,
     else if (fname.extension() == ".npy") {
         // file_impl = new NumpyFile(fname, mode, cfg);
         file_impl = std::make_unique<NumpyFile>(fname, mode, cfg);
+    }else if(fname.extension() == ".dat"){
+        file_impl = std::make_unique<JungfrauDataFile>(fname);
     } else {
         throw std::runtime_error("Unsupported file type");
     }
@@ -73,7 +76,7 @@ size_t File::tell() const { return file_impl->tell(); }
 size_t File::rows() const { return file_impl->rows(); }
 size_t File::cols() const { return file_impl->cols(); }
 size_t File::bitdepth() const { return file_impl->bitdepth(); }
-size_t File::bytes_per_pixel() const { return file_impl->bitdepth() / 8; }
+size_t File::bytes_per_pixel() const { return file_impl->bitdepth() / bits_per_byte; }
 
 DetectorType File::detector_type() const { return file_impl->detector_type(); }
 

src/FilePtr.cpp

@@ -0,0 +1,44 @@
+
+#include "aare/FilePtr.hpp"
+#include <fmt/format.h>
+#include <stdexcept>
+#include <utility>
+
+namespace aare {
+
+FilePtr::FilePtr(const std::filesystem::path& fname, const std::string& mode = "rb") {
+    fp_ = fopen(fname.c_str(), mode.c_str());
+    if (!fp_)
+        throw std::runtime_error(fmt::format("Could not open: {}", fname.c_str()));
+}
+
+FilePtr::FilePtr(FilePtr &&other) { std::swap(fp_, other.fp_); }
+
+FilePtr &FilePtr::operator=(FilePtr &&other) {
+    std::swap(fp_, other.fp_);
+    return *this;
+}
+
+FILE *FilePtr::get() { return fp_; }
+
+int64_t FilePtr::tell() {
+    auto pos = ftell(fp_);
+    if (pos == -1)
+        throw std::runtime_error(fmt::format("Error getting file position: {}", error_msg()));
+    return pos;
+} 
+FilePtr::~FilePtr() {
+    if (fp_)
+        fclose(fp_); // check?
+}
+
+std::string FilePtr::error_msg(){
+    if (feof(fp_)) {
+        return "End of file reached";
+    }
+    if (ferror(fp_)) {
+        return fmt::format("Error reading file: {}", std::strerror(errno));
+    }
+    return "";
+}
+} // namespace aare

src/Fit.cpp

@@ -1,11 +1,13 @@
 #include "aare/Fit.hpp"
 #include "aare/utils/task.hpp"
-
+#include "aare/utils/par.hpp"
 #include <lmcurve2.h>
 #include <lmfit.hpp>
-
 #include <thread>
 
+#include <array>
+
+
 namespace aare {
 
 namespace func {
@@ -16,7 +18,7 @@ double gaus(const double x, const double *par) {
 
 NDArray<double, 1> gaus(NDView<double, 1> x, NDView<double, 1> par) {
     NDArray<double, 1> y({x.shape(0)}, 0);
-    for (size_t i = 0; i < x.size(); i++) {
+    for (ssize_t i = 0; i < x.size(); i++) {
         y(i) = gaus(x(i), par.data());
     }
     return y;
@@ -26,7 +28,7 @@ double pol1(const double x, const double *par) { return par[0] * x + par[1]; }
 
 NDArray<double, 1> pol1(NDView<double, 1> x, NDView<double, 1> par) {
     NDArray<double, 1> y({x.shape()}, 0);
-    for (size_t i = 0; i < x.size(); i++) {
+    for (ssize_t i = 0; i < x.size(); i++) {
         y(i) = pol1(x(i), par.data());
     }
     return y;
@@ -35,33 +37,11 @@ NDArray<double, 1> pol1(NDView<double, 1> x, NDView<double, 1> par) {
 } // namespace func
 
 NDArray<double, 1> fit_gaus(NDView<double, 1> x, NDView<double, 1> y) {
-    NDArray<double, 1> result({3}, 0);
-    lm_control_struct control = lm_control_double;
+    NDArray<double, 1> result = gaus_init_par(x, y);
+    lm_status_struct status;
 
-    // Estimate the initial parameters for the fit
-    std::vector<double> start_par{0, 0, 0};
-    auto e = std::max_element(y.begin(), y.end());
-    auto idx = std::distance(y.begin(), e);
-
-    start_par[0] = *e; // For amplitude we use the maximum value
-    start_par[1] =
-        x[idx]; // For the mean we use the x value of the maximum value
-
-    // For sigma we estimate the fwhm and divide by 2.35
-    // assuming equally spaced x values
-    auto delta = x[1] - x[0];
-    start_par[2] =
-        std::count_if(y.begin(), y.end(),
-                      [e, delta](double val) { return val > *e / 2; }) *
-        delta / 2.35;
-
-    lmfit::result_t res(start_par);
-    lmcurve(res.par.size(), res.par.data(), x.size(), x.data(), y.data(),
-            aare::func::gaus, &control, &res.status);
-
-    result(0) = res.par[0];
-    result(1) = res.par[1];
-    result(2) = res.par[2];
+    lmcurve(result.size(), result.data(), x.size(), x.data(), y.data(),
+            aare::func::gaus, &lm_control_double, &status);
 
     return result;
 }
@@ -81,65 +61,17 @@ NDArray<double, 3> fit_gaus(NDView<double, 1> x, NDView<double, 3> y,
             }
         }
     };
-    auto tasks = split_task(0, y.shape(0), n_threads);
-    std::vector<std::thread> threads;
-    for (auto &task : tasks) {
-        threads.push_back(std::thread(process, task.first, task.second));
-    }
-    for (auto &thread : threads) {
-        thread.join();
-    }
 
+    auto tasks = split_task(0, y.shape(0), n_threads);
+    RunInParallel(process, tasks);
     return result;
 }
 
-void fit_gaus(NDView<double, 1> x, NDView<double, 3> y, NDView<double, 3> y_err,
-              NDView<double, 3> par_out, NDView<double, 3> par_err_out,
-              int n_threads) {
-
-    auto process = [&](ssize_t first_row, ssize_t last_row) {
-        for (ssize_t row = first_row; row < last_row; row++) {
-            for (ssize_t col = 0; col < y.shape(1); col++) {
-                NDView<double, 1> y_view(&y(row, col, 0), {y.shape(2)});
-                NDView<double, 1> y_err_view(&y_err(row, col, 0),
-                                             {y_err.shape(2)});
-                NDView<double, 1> par_out_view(&par_out(row, col, 0),
-                                               {par_out.shape(2)});
-                NDView<double, 1> par_err_out_view(&par_err_out(row, col, 0),
-                                                   {par_err_out.shape(2)});
-                fit_gaus(x, y_view, y_err_view, par_out_view, par_err_out_view);
-            }
-        }
-    };
-
-    auto tasks = split_task(0, y.shape(0), n_threads);
-    std::vector<std::thread> threads;
-    for (auto &task : tasks) {
-        threads.push_back(std::thread(process, task.first, task.second));
-    }
-    for (auto &thread : threads) {
-        thread.join();
-    }
-}
-
-void fit_gaus(NDView<double, 1> x, NDView<double, 1> y, NDView<double, 1> y_err,
-              NDView<double, 1> par_out, NDView<double, 1> par_err_out) {
-    // Check that we have the correct sizes
-    if (y.size() != x.size() || y.size() != y_err.size() ||
-        par_out.size() != 3 || par_err_out.size() != 3) {
-        throw std::runtime_error("Data, x, data_err must have the same size "
-                                 "and par_out, par_err_out must have size 3");
-    }
-
-    lm_control_struct control = lm_control_double;
-
-    // Estimate the initial parameters for the fit
-    std::vector<double> start_par{0, 0, 0};
-    std::vector<double> start_par_err{0, 0, 0};
-    std::vector<double> start_cov{0, 0, 0, 0, 0, 0, 0, 0, 0};
-
+std::array<double, 3> gaus_init_par(const NDView<double, 1> x, const NDView<double, 1> y) {
+    std::array<double, 3> start_par{0, 0, 0};
     auto e = std::max_element(y.begin(), y.end());
     auto idx = std::distance(y.begin(), e);
+
     start_par[0] = *e; // For amplitude we use the maximum value
     start_par[1] =
         x[idx]; // For the mean we use the x value of the maximum value
@@ -152,66 +84,83 @@ void fit_gaus(NDView<double, 1> x, NDView<double, 1> y, NDView<double, 1> y_err,
                       [e, delta](double val) { return val > *e / 2; }) *
         delta / 2.35;
 
-    lmfit::result_t res(start_par);
-    lmfit::result_t res_err(start_par_err);
-    lmfit::result_t cov(start_cov);
-
-    // TODO can we make lmcurve write the result directly where is should be?
-    lmcurve2(res.par.size(), res.par.data(), res_err.par.data(), cov.par.data(),
-             x.size(), x.data(), y.data(), y_err.data(), aare::func::gaus,
-             &control, &res.status);
-
-    par_out(0) = res.par[0];
-    par_out(1) = res.par[1];
-    par_out(2) = res.par[2];
-    par_err_out(0) = res_err.par[0];
-    par_err_out(1) = res_err.par[1];
-    par_err_out(2) = res_err.par[2];
+    return start_par;
 }
 
-void fit_pol1(NDView<double, 1> x, NDView<double, 1> y, NDView<double, 1> y_err,
-              NDView<double, 1> par_out, NDView<double, 1> par_err_out) {
+
+std::array<double, 2> pol1_init_par(const NDView<double, 1> x, const NDView<double, 1> y){
+        // Estimate the initial parameters for the fit
+        std::array<double, 2> start_par{0, 0};
+
+    
+        auto y2 = std::max_element(y.begin(), y.end());
+        auto x2 = x[std::distance(y.begin(), y2)];
+        auto y1 = std::min_element(y.begin(), y.end());
+        auto x1 = x[std::distance(y.begin(), y1)];
+    
+        start_par[0] =
+            (*y2 - *y1) / (x2 - x1); // For amplitude we use the maximum value
+        start_par[1] =
+            *y1 - ((*y2 - *y1) / (x2 - x1)) *
+                      x1; // For the mean we use the x value of the maximum value
+        return start_par;
+}
+
+void fit_gaus(NDView<double, 1> x, NDView<double, 1> y, NDView<double, 1> y_err,
+              NDView<double, 1> par_out, NDView<double, 1> par_err_out,
+              double &chi2) {
+
     // Check that we have the correct sizes
     if (y.size() != x.size() || y.size() != y_err.size() ||
-        par_out.size() != 2 || par_err_out.size() != 2) {
+        par_out.size() != 3 || par_err_out.size() != 3) {
         throw std::runtime_error("Data, x, data_err must have the same size "
-                                 "and par_out, par_err_out must have size 2");
+                                 "and par_out, par_err_out must have size 3");
     }
 
-    lm_control_struct control = lm_control_double;
 
-    // Estimate the initial parameters for the fit
-    std::vector<double> start_par{0, 0};
-    std::vector<double> start_par_err{0, 0};
-    std::vector<double> start_cov{0, 0, 0, 0};
+    // /* Collection of output parameters for status info. */
+    // typedef struct {
+    //     double fnorm;  /* norm of the residue vector fvec. */
+    //     int nfev;      /* actual number of iterations. */
+    //     int outcome;   /* Status indicator. Nonnegative values are used as
+    //     index
+    //                     for the message text lm_infmsg, set in lmmin.c. */
+    //     int userbreak; /* Set when function evaluation requests termination.
+    //     */
+    // } lm_status_struct;
 
-    auto y2 = std::max_element(y.begin(), y.end());
-    auto x2 = x[std::distance(y.begin(), y2)];
-    auto y1 = std::min_element(y.begin(), y.end());
-    auto x1 = x[std::distance(y.begin(), y1)];
 
-    start_par[0] =
-        (*y2 - *y1) / (x2 - x1); // For amplitude we use the maximum value
-    start_par[1] =
-        *y1 - ((*y2 - *y1) / (x2 - x1)) *
-                  x1; // For the mean we use the x value of the maximum value
+    lm_status_struct status;
+    par_out = gaus_init_par(x, y);
+    std::array<double, 9> cov{0, 0, 0, 0, 0, 0, 0 , 0 , 0};
 
-    lmfit::result_t res(start_par);
-    lmfit::result_t res_err(start_par_err);
-    lmfit::result_t cov(start_cov);
+    // void lmcurve2( const int n_par, double *par, double *parerr, double *covar, const int m_dat, const double *t, const double *y, const double *dy, double (*f)( const double ti, const double *par ), const lm_control_struct *control, lm_status_struct *status);
+    // n_par - Number of free variables. Length of parameter vector par.
+    // par - Parameter vector. On input, it must contain a reasonable guess. On output, it contains the solution found to minimize ||r||.
+    // parerr - Parameter uncertainties vector. Array of length n_par or NULL. On output, unless it or covar is NULL, it contains the weighted parameter uncertainties for the found parameters.
+    // covar - Covariance matrix. Array of length n_par * n_par or NULL. On output, unless it is NULL, it contains the covariance matrix.
+    // m_dat - Number of data points. Length of vectors t, y, dy. Must statisfy n_par <= m_dat.
+    // t - Array of length m_dat. Contains the abcissae (time, or "x") for which function f will be evaluated.
+    // y - Array of length m_dat. Contains the ordinate values that shall be fitted.
+    // dy - Array of length m_dat. Contains the standard deviations of the values y.
+    // f - A user-supplied parametric function f(ti;par).
+    // control - Parameter collection for tuning the fit procedure. In most cases, the default &lm_control_double is adequate. If f is only computed with single-precision accuracy, &lm_control_float should be used. Parameters are explained in lmmin2(3).
+    // status - A record used to return information about the minimization process: For details, see lmmin2(3).
 
-    lmcurve2(res.par.size(), res.par.data(), res_err.par.data(), cov.par.data(),
-             x.size(), x.data(), y.data(), y_err.data(), aare::func::pol1,
-             &control, &res.status);
+    lmcurve2(par_out.size(), par_out.data(), par_err_out.data(), cov.data(),
+             x.size(), x.data(), y.data(), y_err.data(), aare::func::gaus,
+             &lm_control_double, &status);
 
-    par_out(0) = res.par[0];
-    par_out(1) = res.par[1];
-    par_err_out(0) = res_err.par[0];
-    par_err_out(1) = res_err.par[1];
+    // Calculate chi2
+    chi2 = 0;
+    for (ssize_t i = 0; i < y.size(); i++) {
+        chi2 += std::pow((y(i) - func::gaus(x(i), par_out.data())) / y_err(i), 2);
+    }
 }
 
-void fit_pol1(NDView<double, 1> x, NDView<double, 3> y, NDView<double, 3> y_err,
-              NDView<double, 3> par_out, NDView<double, 3> par_err_out,
+void fit_gaus(NDView<double, 1> x, NDView<double, 3> y, NDView<double, 3> y_err,
+              NDView<double, 3> par_out, NDView<double, 3> par_err_out,  NDView<double, 2> chi2_out,
+
               int n_threads) {
 
     auto process = [&](ssize_t first_row, ssize_t last_row) {
@@ -224,21 +173,69 @@ void fit_pol1(NDView<double, 1> x, NDView<double, 3> y, NDView<double, 3> y_err,
                                                {par_out.shape(2)});
                 NDView<double, 1> par_err_out_view(&par_err_out(row, col, 0),
                                                    {par_err_out.shape(2)});
-                fit_pol1(x, y_view, y_err_view, par_out_view, par_err_out_view);
+ 
+                fit_gaus(x, y_view, y_err_view, par_out_view, par_err_out_view,
+                         chi2_out(row, col));
+
             }
         }
     };
 
     auto tasks = split_task(0, y.shape(0), n_threads);
-    std::vector<std::thread> threads;
-    for (auto &task : tasks) {
-        threads.push_back(std::thread(process, task.first, task.second));
+    RunInParallel(process, tasks);
+}
+
+void fit_pol1(NDView<double, 1> x, NDView<double, 1> y, NDView<double, 1> y_err,
+              NDView<double, 1> par_out, NDView<double, 1> par_err_out, double& chi2) {
+
+    // Check that we have the correct sizes
+    if (y.size() != x.size() || y.size() != y_err.size() ||
+        par_out.size() != 2 || par_err_out.size() != 2) {
+        throw std::runtime_error("Data, x, data_err must have the same size "
+                                 "and par_out, par_err_out must have size 2");
     }
-    for (auto &thread : threads) {
-        thread.join();
+
+    lm_status_struct status;
+    par_out = pol1_init_par(x, y);
+    std::array<double, 4> cov{0, 0, 0, 0};
+
+    lmcurve2(par_out.size(), par_out.data(), par_err_out.data(), cov.data(),
+             x.size(), x.data(), y.data(), y_err.data(), aare::func::pol1,
+             &lm_control_double, &status);
+
+    // Calculate chi2
+    chi2 = 0;
+    for (ssize_t i = 0; i < y.size(); i++) {
+        chi2 += std::pow((y(i) - func::pol1(x(i), par_out.data())) / y_err(i), 2);
     }
 }
 
+void fit_pol1(NDView<double, 1> x, NDView<double, 3> y, NDView<double, 3> y_err,
+              NDView<double, 3> par_out, NDView<double, 3> par_err_out, NDView<double, 2> chi2_out,
+              int n_threads) {
+
+    auto process = [&](ssize_t first_row, ssize_t last_row) {
+        for (ssize_t row = first_row; row < last_row; row++) {
+            for (ssize_t col = 0; col < y.shape(1); col++) {
+                NDView<double, 1> y_view(&y(row, col, 0), {y.shape(2)});
+                NDView<double, 1> y_err_view(&y_err(row, col, 0),
+                                             {y_err.shape(2)});
+                NDView<double, 1> par_out_view(&par_out(row, col, 0),
+                                               {par_out.shape(2)});
+                NDView<double, 1> par_err_out_view(&par_err_out(row, col, 0),
+                                                   {par_err_out.shape(2)});
+
+                fit_pol1(x, y_view, y_err_view, par_out_view, par_err_out_view, chi2_out(row, col));
+
+            }
+        }
+    };
+
+    auto tasks = split_task(0, y.shape(0), n_threads);
+    RunInParallel(process, tasks);
+
+}
+
 NDArray<double, 1> fit_pol1(NDView<double, 1> x, NDView<double, 1> y) {
     // // Check that we have the correct sizes
     // if (y.size() != x.size() || y.size() != y_err.size() ||
@@ -246,28 +243,12 @@ NDArray<double, 1> fit_pol1(NDView<double, 1> x, NDView<double, 1> y) {
     // throw std::runtime_error("Data, x, data_err must have the same size "
     //                      "and par_out, par_err_out must have size 2");
     // }
-    NDArray<double, 1> par({2}, 0);
+    NDArray<double, 1> par = pol1_init_par(x, y);
 
-    lm_control_struct control = lm_control_double;
+    lm_status_struct status;
+    lmcurve(par.size(), par.data(), x.size(), x.data(), y.data(),
+            aare::func::pol1, &lm_control_double, &status);
 
-    // Estimate the initial parameters for the fit
-    std::vector<double> start_par{0, 0};
-
-    auto y2 = std::max_element(y.begin(), y.end());
-    auto x2 = x[std::distance(y.begin(), y2)];
-    auto y1 = std::min_element(y.begin(), y.end());
-    auto x1 = x[std::distance(y.begin(), y1)];
-
-    start_par[0] = (*y2 - *y1) / (x2 - x1);
-    start_par[1] = *y1 - ((*y2 - *y1) / (x2 - x1)) * x1;
-
-    lmfit::result_t res(start_par);
-
-    lmcurve(res.par.size(), res.par.data(), x.size(), x.data(), y.data(),
-            aare::func::pol1, &control, &res.status);
-
-    par(0) = res.par[0];
-    par(1) = res.par[1];
     return par;
 }
 
@@ -287,13 +268,8 @@ NDArray<double, 3> fit_pol1(NDView<double, 1> x, NDView<double, 3> y,
     };
 
     auto tasks = split_task(0, y.shape(0), n_threads);
-    std::vector<std::thread> threads;
-    for (auto &task : tasks) {
-        threads.push_back(std::thread(process, task.first, task.second));
-    }
-    for (auto &thread : threads) {
-        thread.join();
-    }
+
+    RunInParallel(process, tasks);
     return result;
 }
 

src/Interpolator.cpp

@@ -0,0 +1,56 @@
+#include "aare/Interpolator.hpp"
+
+namespace aare {
+
+Interpolator::Interpolator(NDView<double, 3> etacube, NDView<double, 1> xbins,
+                           NDView<double, 1> ybins, NDView<double, 1> ebins)
+    : m_ietax(etacube), m_ietay(etacube), m_etabinsx(xbins), m_etabinsy(ybins),
+      m_energy_bins(ebins) {
+    if (etacube.shape(0) != xbins.size() || etacube.shape(1) != ybins.size() ||
+        etacube.shape(2) != ebins.size()) {
+        throw std::invalid_argument(
+            "The shape of the etacube does not match the shape of the bins");
+    }
+
+    // Cumulative sum in the x direction
+    for (ssize_t i = 1; i < m_ietax.shape(0); i++) {
+        for (ssize_t j = 0; j < m_ietax.shape(1); j++) {
+            for (ssize_t k = 0; k < m_ietax.shape(2); k++) {
+                m_ietax(i, j, k) += m_ietax(i - 1, j, k);
+            }
+        }
+    }
+
+    // Normalize by the highest row, if norm less than 1 don't do anything
+    for (ssize_t i = 0; i < m_ietax.shape(0); i++) {
+        for (ssize_t j = 0; j < m_ietax.shape(1); j++) {
+            for (ssize_t k = 0; k < m_ietax.shape(2); k++) {
+                auto val = m_ietax(m_ietax.shape(0) - 1, j, k);
+                double norm = val < 1 ? 1 : val;
+                m_ietax(i, j, k) /= norm;
+            }
+        }
+    }
+
+    // Cumulative sum in the y direction
+    for (ssize_t i = 0; i < m_ietay.shape(0); i++) {
+        for (ssize_t j = 1; j < m_ietay.shape(1); j++) {
+            for (ssize_t k = 0; k < m_ietay.shape(2); k++) {
+                m_ietay(i, j, k) += m_ietay(i, j - 1, k);
+            }
+        }
+    }
+
+    // Normalize by the highest column, if norm less than 1 don't do anything
+    for (ssize_t i = 0; i < m_ietay.shape(0); i++) {
+        for (ssize_t j = 0; j < m_ietay.shape(1); j++) {
+            for (ssize_t k = 0; k < m_ietay.shape(2); k++) {
+                auto val = m_ietay(i, m_ietay.shape(1) - 1, k);
+                double norm = val < 1 ? 1 : val;
+                m_ietay(i, j, k) /= norm;
+            }
+        }
+    }
+}
+
+} // namespace aare

src/JungfrauDataFile.cpp

@@ -0,0 +1,238 @@
+#include "aare/JungfrauDataFile.hpp"
+#include "aare/algorithm.hpp"
+#include "aare/defs.hpp"
+
+#include <cerrno>
+#include <fmt/format.h>
+
+namespace aare {
+
+JungfrauDataFile::JungfrauDataFile(const std::filesystem::path &fname) {
+
+    if (!std::filesystem::exists(fname)) {
+        throw std::runtime_error(LOCATION +
+                                 "File does not exist: " + fname.string());
+    }
+    find_frame_size(fname);
+    parse_fname(fname);
+    scan_files();
+    open_file(m_current_file_index);
+}
+
+
+// FileInterface
+
+Frame JungfrauDataFile::read_frame(){
+    Frame f(rows(), cols(), Dtype::UINT16);
+    read_into(reinterpret_cast<std::byte *>(f.data()), nullptr);
+    return f;
+}
+
+Frame JungfrauDataFile::read_frame(size_t frame_number){
+    seek(frame_number);
+    Frame f(rows(), cols(), Dtype::UINT16);
+    read_into(reinterpret_cast<std::byte *>(f.data()), nullptr);
+    return f;
+}
+
+std::vector<Frame> JungfrauDataFile::read_n(size_t n_frames) {
+    std::vector<Frame> frames;
+    for(size_t i = 0; i < n_frames; ++i){
+        frames.push_back(read_frame());
+    }
+    return frames;
+}
+
+void JungfrauDataFile::read_into(std::byte *image_buf) {
+    read_into(image_buf, nullptr);
+}
+void JungfrauDataFile::read_into(std::byte *image_buf, size_t n_frames) {
+    read_into(image_buf, n_frames, nullptr);
+}   
+
+size_t JungfrauDataFile::frame_number(size_t frame_index) {
+    seek(frame_index);
+    return read_header().framenum;
+}
+
+std::array<ssize_t, 2> JungfrauDataFile::shape() const {
+    return {static_cast<ssize_t>(rows()), static_cast<ssize_t>(cols())};
+}
+
+DetectorType JungfrauDataFile::detector_type() const { return DetectorType::Jungfrau; }
+
+std::string JungfrauDataFile::base_name() const { return m_base_name; }
+
+size_t JungfrauDataFile::bytes_per_frame() { return m_bytes_per_frame; }
+
+size_t JungfrauDataFile::pixels_per_frame() { return m_rows * m_cols; }
+
+size_t JungfrauDataFile::bytes_per_pixel() const { return sizeof(pixel_type); }
+
+size_t JungfrauDataFile::bitdepth() const {
+    return bytes_per_pixel() * bits_per_byte;
+}
+
+void JungfrauDataFile::seek(size_t frame_index) {
+    if (frame_index >= m_total_frames) {
+        throw std::runtime_error(LOCATION + "Frame index out of range: " +
+                                 std::to_string(frame_index));
+    }
+    m_current_frame_index = frame_index;
+    auto file_index = first_larger(m_last_frame_in_file, frame_index);
+
+    if (file_index != m_current_file_index)
+        open_file(file_index);
+
+    auto frame_offset = (file_index)
+                            ? frame_index - m_last_frame_in_file[file_index - 1]
+                            : frame_index;
+    auto byte_offset = frame_offset * (m_bytes_per_frame + header_size);
+    m_fp.seek(byte_offset);
+};
+
+size_t JungfrauDataFile::tell() { return m_current_frame_index; }
+size_t JungfrauDataFile::total_frames() const { return m_total_frames; }
+size_t JungfrauDataFile::rows() const { return m_rows; }
+size_t JungfrauDataFile::cols() const { return m_cols; }
+
+size_t JungfrauDataFile::n_files() const { return m_last_frame_in_file.size(); }
+
+void JungfrauDataFile::find_frame_size(const std::filesystem::path &fname) {
+
+    static constexpr size_t module_data_size =
+        header_size + sizeof(pixel_type) * 512 * 1024;
+    static constexpr size_t half_data_size =
+        header_size + sizeof(pixel_type) * 256 * 1024;
+    static constexpr size_t chip_data_size =
+        header_size + sizeof(pixel_type) * 256 * 256;
+
+    auto file_size = std::filesystem::file_size(fname);
+    if (file_size == 0) {
+        throw std::runtime_error(LOCATION +
+                                 "Cannot guess frame size: file is empty");
+    }
+
+    if (file_size % module_data_size == 0) {
+        m_rows = 512;
+        m_cols = 1024;
+        m_bytes_per_frame = module_data_size - header_size;
+    } else if (file_size % half_data_size == 0) {
+        m_rows = 256;
+        m_cols = 1024;
+        m_bytes_per_frame = half_data_size - header_size;
+    } else if (file_size % chip_data_size == 0) {
+        m_rows = 256;
+        m_cols = 256;
+        m_bytes_per_frame = chip_data_size - header_size;
+    } else {
+        throw std::runtime_error(LOCATION +
+                                 "Cannot find frame size: file size is not a "
+                                 "multiple of any known frame size");
+    }
+}
+
+void JungfrauDataFile::parse_fname(const std::filesystem::path &fname) {
+    m_path = fname.parent_path();
+    m_base_name = fname.stem();
+
+    // find file index, then remove if from the base name
+    if (auto pos = m_base_name.find_last_of('_'); pos != std::string::npos) {
+        m_offset = std::stoul(m_base_name.substr(pos + 1));
+        m_base_name.erase(pos);
+    }
+}
+
+void JungfrauDataFile::scan_files() {
+    // find how many files we have and the number of frames in each file
+    m_last_frame_in_file.clear();
+    size_t file_index = m_offset;
+    while (std::filesystem::exists(fpath(file_index))) {
+        auto n_frames = std::filesystem::file_size(fpath(file_index)) /
+                        (m_bytes_per_frame + header_size);
+        m_last_frame_in_file.push_back(n_frames);
+        ++file_index;
+    }
+
+    // find where we need to open the next file and total number of frames
+    m_last_frame_in_file = cumsum(m_last_frame_in_file);
+    m_total_frames = m_last_frame_in_file.back();
+}
+
+void JungfrauDataFile::read_into(std::byte *image_buf,
+                                 JungfrauDataHeader *header) {
+
+    // read header if not passed nullptr
+    if (header) {
+        if (auto rc = fread(header, sizeof(JungfrauDataHeader), 1, m_fp.get());
+            rc != 1) {
+            throw std::runtime_error(
+                LOCATION +
+                "Could not read header from file:" + m_fp.error_msg());
+        }
+    } else {
+        m_fp.seek(header_size, SEEK_CUR);
+    }
+
+    // read data
+    if (auto rc = fread(image_buf, 1, m_bytes_per_frame, m_fp.get());
+        rc != m_bytes_per_frame) {
+        throw std::runtime_error(LOCATION + "Could not read image from file" +
+                                 m_fp.error_msg());
+    }
+
+    // prepare for next read
+    // if we are at the end of the file, open the next file
+    ++m_current_frame_index;
+    if (m_current_frame_index >= m_last_frame_in_file[m_current_file_index] &&
+        (m_current_frame_index < m_total_frames)) {
+        ++m_current_file_index;
+        open_file(m_current_file_index);
+    }
+}
+
+void JungfrauDataFile::read_into(std::byte *image_buf, size_t n_frames,
+                                 JungfrauDataHeader *header) {
+    if (header) {
+        for (size_t i = 0; i < n_frames; ++i)
+            read_into(image_buf + i * m_bytes_per_frame, header + i); 
+    }else{
+        for (size_t i = 0; i < n_frames; ++i)
+            read_into(image_buf + i * m_bytes_per_frame, nullptr);
+    }
+}
+
+void JungfrauDataFile::read_into(NDArray<uint16_t>* image, JungfrauDataHeader* header) {
+    if(image->shape()!=shape()){
+        throw std::runtime_error(LOCATION +
+            "Image shape does not match file size: " + std::to_string(rows()) + "x" + std::to_string(cols()));
+    }
+    read_into(reinterpret_cast<std::byte *>(image->data()), header);
+}
+
+
+JungfrauDataHeader JungfrauDataFile::read_header() {
+    JungfrauDataHeader header;
+    if (auto rc = fread(&header, 1, sizeof(header), m_fp.get());
+        rc != sizeof(header)) {
+        throw std::runtime_error(LOCATION + "Could not read header from file" +
+                                 m_fp.error_msg());
+    }
+    m_fp.seek(-header_size, SEEK_CUR);
+    return header;
+}
+
+void JungfrauDataFile::open_file(size_t file_index) {
+    // fmt::print(stderr, "Opening file: {}\n",
+    // fpath(file_index+m_offset).string());
+    m_fp = FilePtr(fpath(file_index + m_offset), "rb");
+    m_current_file_index = file_index;
+}
+
+std::filesystem::path JungfrauDataFile::fpath(size_t file_index) const {
+    auto fname = fmt::format("{}_{:0{}}.dat", m_base_name, file_index,
+                             n_digits_in_file_index);
+    return m_path / fname;
+}
+
+} // namespace aare

src/JungfrauDataFile.test.cpp

@@ -0,0 +1,114 @@
+#include "aare/JungfrauDataFile.hpp"
+
+#include <catch2/catch_test_macros.hpp>
+#include "test_config.hpp"
+
+using aare::JungfrauDataFile;
+using aare::JungfrauDataHeader;
+TEST_CASE("Open a Jungfrau data file", "[.files]") {
+    //we know we have 4 files with 7, 7, 7, and 3 frames
+    //firs frame number if 1 and the bunch id is frame_number**2
+    //so we can check the header
+    auto fpath = test_data_path() / "dat" / "AldoJF500k_000000.dat";
+    REQUIRE(std::filesystem::exists(fpath));
+
+    JungfrauDataFile f(fpath);
+    REQUIRE(f.rows() == 512);
+    REQUIRE(f.cols() == 1024);
+    REQUIRE(f.bytes_per_frame() == 1048576);    
+    REQUIRE(f.pixels_per_frame() == 524288);
+    REQUIRE(f.bytes_per_pixel() == 2);
+    REQUIRE(f.bitdepth() == 16);
+    REQUIRE(f.base_name() == "AldoJF500k");
+    REQUIRE(f.n_files() == 4);
+    REQUIRE(f.tell() == 0);
+    REQUIRE(f.total_frames() == 24);
+    REQUIRE(f.current_file() == fpath);
+
+    //Check that the frame number and buch id is read correctly
+    for (size_t i = 0; i < 24; ++i) {
+        JungfrauDataHeader header;
+        aare::NDArray<uint16_t> image(f.shape());
+        f.read_into(&image, &header);
+        REQUIRE(header.framenum == i + 1);
+        REQUIRE(header.bunchid == (i + 1) * (i + 1));
+        REQUIRE(image.shape(0) == 512);
+        REQUIRE(image.shape(1) == 1024);
+    }
+}
+
+TEST_CASE("Seek in a JungfrauDataFile", "[.files]"){
+    auto fpath = test_data_path() / "dat" / "AldoJF65k_000000.dat";
+    REQUIRE(std::filesystem::exists(fpath));
+
+    JungfrauDataFile f(fpath);
+
+    //The file should have 113 frames
+    f.seek(19);
+    REQUIRE(f.tell() == 19);
+    auto h = f.read_header();
+    REQUIRE(h.framenum == 19+1);
+
+    //Reading again does not change the file pointer
+    auto h2 = f.read_header();
+    REQUIRE(h2.framenum == 19+1);
+
+    f.seek(59);
+    REQUIRE(f.tell() == 59);
+    auto h3 = f.read_header();
+    REQUIRE(h3.framenum == 59+1);
+
+    JungfrauDataHeader h4;
+    aare::NDArray<uint16_t> image(f.shape());
+    f.read_into(&image, &h4);
+    REQUIRE(h4.framenum == 59+1);
+
+    //now we should be on the next frame
+    REQUIRE(f.tell() == 60);
+    REQUIRE(f.read_header().framenum == 60+1);
+
+    REQUIRE_THROWS(f.seek(86356)); //out of range
+}
+
+TEST_CASE("Open a Jungfrau data file with non zero file index", "[.files]"){
+
+    auto fpath = test_data_path() / "dat" / "AldoJF65k_000003.dat";
+    REQUIRE(std::filesystem::exists(fpath));
+
+    JungfrauDataFile f(fpath);
+
+    //18 files per data file, opening the 3rd file we ignore the first 3
+    REQUIRE(f.total_frames() == 113-18*3);
+    REQUIRE(f.tell() == 0);
+
+    //Frame numbers start at 1 in the first file
+    REQUIRE(f.read_header().framenum == 18*3+1);
+
+    // moving relative to the third file
+    f.seek(5);
+    REQUIRE(f.read_header().framenum == 18*3+1+5);
+
+    // ignoring the first 3 files
+    REQUIRE(f.n_files() == 4);
+
+    REQUIRE(f.current_file().stem() == "AldoJF65k_000003");
+
+}
+
+TEST_CASE("Read into throws if size doesn't match", "[.files]"){
+    auto fpath = test_data_path() / "dat" / "AldoJF65k_000000.dat";
+    REQUIRE(std::filesystem::exists(fpath));
+
+    JungfrauDataFile f(fpath);
+
+    aare::NDArray<uint16_t> image({39, 85});
+    JungfrauDataHeader header;
+
+    REQUIRE_THROWS(f.read_into(&image, &header));
+    REQUIRE_THROWS(f.read_into(&image, nullptr));
+    REQUIRE_THROWS(f.read_into(&image));
+
+    REQUIRE(f.tell() == 0);
+
+
+}

src/NDArray.test.cpp

@@ -2,6 +2,7 @@
 #include <array>
 #include <catch2/benchmark/catch_benchmark.hpp>
 #include <catch2/catch_test_macros.hpp>
+#include <numeric>
 
 using aare::NDArray;
 using aare::NDView;
@@ -34,6 +35,24 @@ TEST_CASE("Construct from an NDView") {
     }
 }
 
+TEST_CASE("3D NDArray from NDView"){
+    std::vector<int> data(27);
+    std::iota(data.begin(), data.end(), 0);
+    NDView<int, 3> view(data.data(), Shape<3>{3, 3, 3});
+    NDArray<int, 3> image(view);
+    REQUIRE(image.shape() == view.shape());
+    REQUIRE(image.size() == view.size());
+    REQUIRE(image.data() != view.data());
+
+    for(int64_t i=0; i<image.shape(0); i++){
+        for(int64_t j=0; j<image.shape(1); j++){
+            for(int64_t k=0; k<image.shape(2); k++){
+                REQUIRE(image(i, j, k) == view(i, j, k));
+            }
+        }
+    }
+}
+
 TEST_CASE("1D image") {
     std::array<int64_t, 1> shape{{20}};
     NDArray<short, 1> img(shape, 3);
@@ -164,14 +183,14 @@ TEST_CASE("Size and shape matches") {
     int64_t h = 75;
     std::array<int64_t, 2> shape{w, h};
     NDArray<double> a{shape};
-    REQUIRE(a.size() == static_cast<uint64_t>(w * h));
+    REQUIRE(a.size() == w * h);
     REQUIRE(a.shape() == shape);
 }
 
 TEST_CASE("Initial value matches for all elements") {
     double v = 4.35;
     NDArray<double> a{{5, 5}, v};
-    for (uint32_t i = 0; i < a.size(); ++i) {
+    for (int i = 0; i < a.size(); ++i) {
         REQUIRE(a(i) == v);
     }
 }
@@ -379,4 +398,32 @@ TEST_CASE("Elementwise operations on images") {
             REQUIRE(A(i) == a_val);
         }
     }
+}
+
+TEST_CASE("Assign an std::array to a 1D NDArray") {
+    NDArray<int, 1> a{{5}, 0};
+    std::array<int, 5> b{1, 2, 3, 4, 5};
+    a = b;
+    for (uint32_t i = 0; i < a.size(); ++i) {
+        REQUIRE(a(i) == b[i]);
+    }
+}
+
+TEST_CASE("Assign an std::array to a 1D NDArray of a different size") {
+    NDArray<int, 1> a{{3}, 0};
+    std::array<int, 5> b{1, 2, 3, 4, 5};
+    a = b;
+
+    REQUIRE(a.size() == 5);
+    for (uint32_t i = 0; i < a.size(); ++i) {
+        REQUIRE(a(i) == b[i]);
+    }
+}
+
+TEST_CASE("Construct an NDArray from an std::array") {
+    std::array<int, 5> b{1, 2, 3, 4, 5};
+    NDArray<int, 1> a(b);
+    for (uint32_t i = 0; i < a.size(); ++i) {
+        REQUIRE(a(i) == b[i]);
+    }
 }

src/NDView.test.cpp

@@ -3,6 +3,7 @@
 
 #include <iostream>
 #include <vector>
+#include <numeric>
 
 using aare::NDView;
 using aare::Shape;
@@ -21,10 +22,8 @@ TEST_CASE("Element reference 1D") {
 }
 
 TEST_CASE("Element reference 2D") {
-    std::vector<int> vec;
-    for (int i = 0; i != 12; ++i) {
-        vec.push_back(i);
-    }
+    std::vector<int> vec(12);
+    std::iota(vec.begin(), vec.end(), 0);
 
     NDView<int, 2> data(vec.data(), Shape<2>{3, 4});
     REQUIRE(vec.size() == static_cast<size_t>(data.size()));
@@ -58,10 +57,8 @@ TEST_CASE("Element reference 3D") {
 }
 
 TEST_CASE("Plus and miuns with single value") {
-    std::vector<int> vec;
-    for (int i = 0; i != 12; ++i) {
-        vec.push_back(i);
-    }
+    std::vector<int> vec(12);
+    std::iota(vec.begin(), vec.end(), 0);
     NDView<int, 2> data(vec.data(), Shape<2>{3, 4});
     data += 5;
     int i = 0;
@@ -116,10 +113,8 @@ TEST_CASE("elementwise assign") {
 }
 
 TEST_CASE("iterators") {
-    std::vector<int> vec;
-    for (int i = 0; i != 12; ++i) {
-        vec.push_back(i);
-    }
+    std::vector<int> vec(12);
+    std::iota(vec.begin(), vec.end(), 0);
     NDView<int, 1> data(vec.data(), Shape<1>{12});
     int i = 0;
     for (const auto item : data) {
@@ -167,27 +162,31 @@ TEST_CASE("divide with another span") {
 }
 
 TEST_CASE("Retrieve shape") {
-    std::vector<int> vec;
-    for (int i = 0; i != 12; ++i) {
-        vec.push_back(i);
-    }
+    std::vector<int> vec(12);
+    std::iota(vec.begin(), vec.end(), 0);
     NDView<int, 2> data(vec.data(), Shape<2>{3, 4});
     REQUIRE(data.shape()[0] == 3);
     REQUIRE(data.shape()[1] == 4);
 }
 
 TEST_CASE("compare two views") {
-    std::vector<int> vec1;
-    for (int i = 0; i != 12; ++i) {
-        vec1.push_back(i);
-    }
+    std::vector<int> vec1(12);
+    std::iota(vec1.begin(), vec1.end(), 0);
     NDView<int, 2> view1(vec1.data(), Shape<2>{3, 4});
 
-    std::vector<int> vec2;
-    for (int i = 0; i != 12; ++i) {
-        vec2.push_back(i);
-    }
+    std::vector<int> vec2(12);
+    std::iota(vec2.begin(), vec2.end(), 0);
     NDView<int, 2> view2(vec2.data(), Shape<2>{3, 4});
 
     REQUIRE((view1 == view2));
+}
+
+
+TEST_CASE("Create a view over a vector"){
+    std::vector<int> vec(12);
+    std::iota(vec.begin(), vec.end(), 0);
+    auto v = aare::make_view(vec);
+    REQUIRE(v.shape()[0] == 12);
+    REQUIRE(v[0] == 0);
+    REQUIRE(v[11] == 11);
 }

src/RawFile.cpp

@@ -76,8 +76,7 @@ size_t RawFile::n_mod() const { return n_subfile_parts; }
 
 
 size_t RawFile::bytes_per_frame() {
-    // return m_rows * m_cols * m_master.bitdepth() / 8;
-    return m_geometry.pixels_x * m_geometry.pixels_y * m_master.bitdepth() / 8;
+    return m_geometry.pixels_x * m_geometry.pixels_y * m_master.bitdepth() / bits_per_byte;
 }
 size_t RawFile::pixels_per_frame() { 
     // return m_rows * m_cols; 

src/RawSubFile.cpp

@@ -1,9 +1,12 @@
 #include "aare/RawSubFile.hpp"
 #include "aare/PixelMap.hpp"
+#include "aare/utils/ifstream_helpers.hpp"
 #include <cstring> // memcpy
 #include <fmt/core.h>
 #include <iostream>
 
+
+
 namespace aare {
 
 RawSubFile::RawSubFile(const std::filesystem::path &fname,
@@ -20,7 +23,7 @@ RawSubFile::RawSubFile(const std::filesystem::path &fname,
     }
 
     if (std::filesystem::exists(fname)) {
-        n_frames = std::filesystem::file_size(fname) /
+        m_num_frames = std::filesystem::file_size(fname) /
                    (sizeof(DetectorHeader) + rows * cols * bitdepth / 8);
     } else {
         throw std::runtime_error(
@@ -35,7 +38,7 @@ RawSubFile::RawSubFile(const std::filesystem::path &fname,
     }
 
 #ifdef AARE_VERBOSE
-    fmt::print("Opened file: {} with {} frames\n", m_fname.string(), n_frames);
+    fmt::print("Opened file: {} with {} frames\n", m_fname.string(), m_num_frames);
     fmt::print("m_rows: {}, m_cols: {}, m_bitdepth: {}\n", m_rows, m_cols,
                m_bitdepth);
     fmt::print("file size: {}\n", std::filesystem::file_size(fname));
@@ -43,8 +46,8 @@ RawSubFile::RawSubFile(const std::filesystem::path &fname,
 }
 
 void RawSubFile::seek(size_t frame_index) {
-    if (frame_index >= n_frames) {
-        throw std::runtime_error(LOCATION + fmt::format("Frame index {} out of range in a file with {} frames", frame_index, n_frames));
+    if (frame_index >= m_num_frames) {
+        throw std::runtime_error(LOCATION + fmt::format("Frame index {} out of range in a file with {} frames", frame_index, m_num_frames));
     }
     m_file.seekg((sizeof(DetectorHeader) + bytes_per_frame()) * frame_index);
 }
@@ -60,6 +63,10 @@ void RawSubFile::read_into(std::byte *image_buf, DetectorHeader *header) {
         m_file.seekg(sizeof(DetectorHeader), std::ios::cur);
     }
 
+    if (m_file.fail()){
+        throw std::runtime_error(LOCATION + ifstream_error_msg(m_file));
+    }
+
     // TODO! expand support for different bitdepths
     if (m_pixel_map) {
         // read into a temporary buffer and then copy the data to the buffer
@@ -79,8 +86,24 @@ void RawSubFile::read_into(std::byte *image_buf, DetectorHeader *header) {
         // read directly into the buffer
         m_file.read(reinterpret_cast<char *>(image_buf), bytes_per_frame());
     }
+
+    if (m_file.fail()){
+        throw std::runtime_error(LOCATION + ifstream_error_msg(m_file));
+    }
 }
 
+void RawSubFile::read_into(std::byte *image_buf, size_t n_frames, DetectorHeader *header) {
+    for (size_t i = 0; i < n_frames; i++) {
+        read_into(image_buf, header);
+        image_buf += bytes_per_frame();
+        if (header) {
+            ++header;
+        }
+    }
+}
+
+
+
 template <typename T>
 void RawSubFile::read_with_map(std::byte *image_buf) {
     auto part_buffer = new std::byte[bytes_per_frame()];

src/algorithm.test.cpp

@@ -0,0 +1,162 @@
+
+
+#include <aare/algorithm.hpp>
+#include <catch2/catch_test_macros.hpp>
+
+TEST_CASE("Find the closed index in a 1D array", "[algorithm]") {
+    aare::NDArray<double, 1> arr({5});
+    for (ssize_t i = 0; i < arr.size(); i++) {
+        arr[i] = i;
+    }
+    // arr 0, 1, 2, 3, 4
+    REQUIRE(aare::nearest_index(arr, 2.3) == 2);
+    REQUIRE(aare::nearest_index(arr, 2.6) == 3);
+    REQUIRE(aare::nearest_index(arr, 45.0) == 4);
+    REQUIRE(aare::nearest_index(arr, 0.0) == 0);
+    REQUIRE(aare::nearest_index(arr, -1.0) == 0);
+}
+
+TEST_CASE("Passing integers to nearest_index works", "[algorithm]") {
+    aare::NDArray<int, 1> arr({5});
+    for (ssize_t i = 0; i < arr.size(); i++) {
+        arr[i] = i;
+    }
+    // arr 0, 1, 2, 3, 4
+    REQUIRE(aare::nearest_index(arr, 2) == 2);
+    REQUIRE(aare::nearest_index(arr, 3) == 3);
+    REQUIRE(aare::nearest_index(arr, 45) == 4);
+    REQUIRE(aare::nearest_index(arr, 0) == 0);
+    REQUIRE(aare::nearest_index(arr, -1) == 0);
+}
+
+TEST_CASE("nearest_index works with std::vector", "[algorithm]") {
+    std::vector<double> vec = {0, 1, 2, 3, 4};
+    REQUIRE(aare::nearest_index(vec, 2.123) == 2);
+    REQUIRE(aare::nearest_index(vec, 2.66) == 3);
+    REQUIRE(aare::nearest_index(vec, 4555555.0) == 4);
+    REQUIRE(aare::nearest_index(vec, 0.0) == 0);
+    REQUIRE(aare::nearest_index(vec, -10.0) == 0);
+}
+
+TEST_CASE("nearest index works with std::array", "[algorithm]") {
+    std::array<double, 5> arr = {0, 1, 2, 3, 4};
+    REQUIRE(aare::nearest_index(arr, 2.123) == 2);
+    REQUIRE(aare::nearest_index(arr, 2.501) == 3);
+    REQUIRE(aare::nearest_index(arr, 4555555.0) == 4);
+    REQUIRE(aare::nearest_index(arr, 0.0) == 0);
+    REQUIRE(aare::nearest_index(arr, -10.0) == 0);
+}
+
+TEST_CASE("nearest index when there is no different uses the first element",
+          "[algorithm]") {
+    std::vector<int> vec = {5, 5, 5, 5, 5};
+    REQUIRE(aare::nearest_index(vec, 5) == 0);
+}
+
+TEST_CASE("nearest index when there is no different uses the first element "
+          "also when all smaller",
+          "[algorithm]") {
+    std::vector<int> vec = {5, 5, 5, 5, 5};
+    REQUIRE(aare::nearest_index(vec, 10) == 0);
+}
+
+TEST_CASE("last smaller", "[algorithm]") {
+    aare::NDArray<double, 1> arr({5});
+    for (ssize_t i = 0; i < arr.size(); i++) {
+        arr[i] = i;
+    }
+    // arr 0, 1, 2, 3, 4
+    REQUIRE(aare::last_smaller(arr, -10.0) == 0);
+    REQUIRE(aare::last_smaller(arr, 0.0) == 0);
+    REQUIRE(aare::last_smaller(arr, 2.3) == 2);
+    REQUIRE(aare::last_smaller(arr, 253.) == 4);
+}
+
+TEST_CASE("returns last bin strictly smaller", "[algorithm]") {
+    aare::NDArray<double, 1> arr({5});
+    for (ssize_t i = 0; i < arr.size(); i++) {
+        arr[i] = i;
+    }
+    // arr 0, 1, 2, 3, 4
+    REQUIRE(aare::last_smaller(arr, 2.0) == 1);
+}
+
+TEST_CASE("last_smaller with all elements smaller returns last element",
+          "[algorithm]") {
+    aare::NDArray<double, 1> arr({5});
+    for (ssize_t i = 0; i < arr.size(); i++) {
+        arr[i] = i;
+    }
+    // arr 0, 1, 2, 3, 4
+    REQUIRE(aare::last_smaller(arr, 50.) == 4);
+}
+
+TEST_CASE("last_smaller with all elements bigger returns first element",
+          "[algorithm]") {
+    aare::NDArray<double, 1> arr({5});
+    for (ssize_t i = 0; i < arr.size(); i++) {
+        arr[i] = i;
+    }
+    // arr 0, 1, 2, 3, 4
+    REQUIRE(aare::last_smaller(arr, -50.) == 0);
+}
+
+TEST_CASE("last smaller with all elements equal returns the first element",
+          "[algorithm]") {
+    std::vector<int> vec = {5, 5, 5, 5, 5, 5, 5};
+    REQUIRE(aare::last_smaller(vec, 5) == 0);
+}
+
+TEST_CASE("first_lager with vector", "[algorithm]") {
+    std::vector<double> vec = {0, 1, 2, 3, 4};
+    REQUIRE(aare::first_larger(vec, 2.5) == 3);
+}
+
+TEST_CASE("first_lager with all elements smaller returns last element",
+          "[algorithm]") {
+    std::vector<double> vec = {0, 1, 2, 3, 4};
+    REQUIRE(aare::first_larger(vec, 50.) == 4);
+}
+
+TEST_CASE("first_lager with all elements bigger returns first element",
+          "[algorithm]") {
+    std::vector<double> vec = {0, 1, 2, 3, 4};
+    REQUIRE(aare::first_larger(vec, -50.) == 0);
+}
+
+TEST_CASE("first_lager with all elements the same as the check returns last",
+          "[algorithm]") {
+    std::vector<int> vec = {14, 14, 14, 14, 14};
+    REQUIRE(aare::first_larger(vec, 14) == 4);
+}
+
+TEST_CASE("first larger with the same element", "[algorithm]") {
+    std::vector<int> vec = {7, 8, 9, 10, 11};
+    REQUIRE(aare::first_larger(vec, 9) == 3);
+}
+
+TEST_CASE("cumsum works", "[algorithm]") {
+    std::vector<double> vec = {0, 1, 2, 3, 4};
+    auto result = aare::cumsum(vec);
+    REQUIRE(result.size() == vec.size());
+    REQUIRE(result[0] == 0);
+    REQUIRE(result[1] == 1);
+    REQUIRE(result[2] == 3);
+    REQUIRE(result[3] == 6);
+    REQUIRE(result[4] == 10);
+}
+TEST_CASE("cumsum works with empty vector", "[algorithm]") {
+    std::vector<double> vec = {};
+    auto result = aare::cumsum(vec);
+    REQUIRE(result.size() == 0);
+}
+TEST_CASE("cumsum works with negative numbers", "[algorithm]") {
+    std::vector<double> vec = {0, -1, -2, -3, -4};
+    auto result = aare::cumsum(vec);
+    REQUIRE(result.size() == vec.size());
+    REQUIRE(result[0] == 0);
+    REQUIRE(result[1] == -1);
+    REQUIRE(result[2] == -3);
+    REQUIRE(result[3] == -6);
+    REQUIRE(result[4] == -10);
+}

src/decode.cpp

@@ -1,5 +1,5 @@
 #include "aare/decode.hpp"
-
+#include <cmath>
 namespace aare {
 
 uint16_t adc_sar_05_decode64to16(uint64_t input){
@@ -22,6 +22,10 @@ uint16_t adc_sar_05_decode64to16(uint64_t input){
 }
 
 void adc_sar_05_decode64to16(NDView<uint64_t, 2> input, NDView<uint16_t,2> output){
+    if(input.shape() != output.shape()){
+        throw std::invalid_argument(LOCATION + " input and output shapes must match");
+    }
+
     for(int64_t i = 0; i < input.shape(0); i++){
         for(int64_t j = 0; j < input.shape(1); j++){
             output(i,j) = adc_sar_05_decode64to16(input(i,j));
@@ -49,6 +53,9 @@ uint16_t adc_sar_04_decode64to16(uint64_t input){
 }
 
 void adc_sar_04_decode64to16(NDView<uint64_t, 2> input, NDView<uint16_t,2> output){
+    if(input.shape() != output.shape()){
+        throw std::invalid_argument(LOCATION + " input and output shapes must match");
+    }
     for(int64_t i = 0; i < input.shape(0); i++){
         for(int64_t j = 0; j < input.shape(1); j++){
             output(i,j) = adc_sar_04_decode64to16(input(i,j));
@@ -56,6 +63,40 @@ void adc_sar_04_decode64to16(NDView<uint64_t, 2> input, NDView<uint16_t,2> outpu
     }
 }
 
+double apply_custom_weights(uint16_t input, const NDView<double, 1> weights) {
+    if(weights.size() > 16){
+        throw std::invalid_argument("weights size must be less than or equal to 16");
+    }
+
+    double result = 0.0;
+    for (ssize_t i = 0; i < weights.size(); ++i) {
+        result += ((input >> i) & 1) * std::pow(weights[i], i);
+    }
+    return result;
+
+}
+
+void apply_custom_weights(NDView<uint16_t, 1> input, NDView<double, 1> output, const NDView<double,1> weights) {
+    if(input.shape() != output.shape()){
+        throw std::invalid_argument(LOCATION + " input and output shapes must match");
+    }
+
+    //Calculate weights to avoid repeatedly calling std::pow
+    std::vector<double> weights_powers(weights.size());
+    for (ssize_t i = 0; i < weights.size(); ++i) {
+        weights_powers[i] = std::pow(weights[i], i);
+    }
+
+    // Apply custom weights to each element in the input array
+    for (ssize_t i = 0; i < input.shape(0); i++) {            
+        double result = 0.0;
+        for (size_t bit_index = 0; bit_index < weights_powers.size(); ++bit_index) {
+            result += ((input(i) >> bit_index) & 1) * weights_powers[bit_index];
+        }
+        output(i) = result;
+    }
+}
+
 
 
 } // namespace aare

src/decode.test.cpp

@@ -0,0 +1,80 @@
+#include "aare/decode.hpp"
+
+#include <catch2/matchers/catch_matchers_floating_point.hpp>
+#include <catch2/catch_test_macros.hpp>
+#include "aare/NDArray.hpp"
+using Catch::Matchers::WithinAbs;
+#include <vector>
+
+TEST_CASE("test_adc_sar_05_decode64to16"){
+    uint64_t input = 0;
+    uint16_t output = aare::adc_sar_05_decode64to16(input);
+    CHECK(output == 0);
+
+
+    // bit 29 on th input is bit 0 on the output
+    input = 1UL << 29;
+    output = aare::adc_sar_05_decode64to16(input);
+    CHECK(output == 1);
+
+    // test all bits by iteratting through the bitlist
+    std::vector<int> bitlist = {29, 19, 28, 18, 31, 21, 27, 20, 24, 23, 25, 22};
+    for (size_t i = 0; i < bitlist.size(); i++) {
+        input = 1UL << bitlist[i];
+        output = aare::adc_sar_05_decode64to16(input);
+        CHECK(output == (1 << i));
+    }
+
+    
+    // test a few "random" values
+    input = 0;
+    input |= (1UL << 29);
+    input |= (1UL << 19);
+    input |= (1UL << 28);
+    output = aare::adc_sar_05_decode64to16(input);
+    CHECK(output == 7UL);
+
+
+    input = 0;
+    input |= (1UL << 18);
+    input |= (1UL << 27);
+    input |= (1UL << 25);
+    output = aare::adc_sar_05_decode64to16(input);
+    CHECK(output == 1096UL);
+
+    input = 0;
+    input |= (1UL << 25);
+    input |= (1UL << 22);
+    output = aare::adc_sar_05_decode64to16(input);
+    CHECK(output == 3072UL);
+ }
+
+
+ TEST_CASE("test_apply_custom_weights") {
+
+    uint16_t input = 1;
+    aare::NDArray<double, 1> weights_data({3}, 0.0);
+    weights_data(0) = 1.7;
+    weights_data(1) = 2.1;
+    weights_data(2) = 1.8;
+
+    auto weights = weights_data.view();
+
+
+    double output = aare::apply_custom_weights(input, weights);
+    CHECK_THAT(output, WithinAbs(1.0, 0.001));
+
+    input = 1 << 1;
+    output = aare::apply_custom_weights(input, weights);
+    CHECK_THAT(output, WithinAbs(2.1, 0.001));
+
+
+    input = 1 << 2;
+    output = aare::apply_custom_weights(input, weights);
+    CHECK_THAT(output, WithinAbs(3.24, 0.001));
+
+    input = 0b111;
+    output = aare::apply_custom_weights(input, weights);
+    CHECK_THAT(output, WithinAbs(6.34, 0.001));
+
+ }

src/utils/ifstream_helpers.cpp

@@ -0,0 +1,18 @@
+#include "aare/utils/ifstream_helpers.hpp"
+
+namespace aare {
+
+std::string ifstream_error_msg(std::ifstream &ifs) {
+    std::ios_base::iostate state = ifs.rdstate();
+    if (state & std::ios_base::eofbit) {
+        return " End of file reached";
+    } else if (state & std::ios_base::badbit) {
+        return " Bad file stream";
+    } else if (state & std::ios_base::failbit) {
+        return " File read failed";
+    }else{
+        return " Unknown/no error";
+    }
+}
+
+} // namespace aare

tests/test_config.hpp.in

@@ -7,6 +7,6 @@ inline auto test_data_path(){
     if(const char* env_p = std::getenv("AARE_TEST_DATA")){
         return std::filesystem::path(env_p);
     }else{
-        throw std::runtime_error("AARE_TEST_DATA_PATH not set");
+        throw std::runtime_error("Path to test data: $AARE_TEST_DATA not set");
     }
 }