Merge branch 'main' into dev/reduce

- added ClusterVector::empty() to check if the vector is empty
- Fixed generation of missing docs for ClusterVector

.gitea/workflows/rh8-native.yml

@@ -1,18 +1,24 @@
 name: Build on RHEL8
 
 on:
+  push:
   workflow_dispatch:
 
 permissions:
   contents: read
 
 jobs:
-  buildh:
+  build:
     runs-on: "ubuntu-latest"
     container:
       image: gitea.psi.ch/images/rhel8-developer-gitea-actions
     steps:
-      - uses: actions/checkout@v4
+      # 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
@@ -22,7 +28,7 @@ jobs:
       - name: Build library 
         run: |
             mkdir build && cd build
-            cmake .. -DAARE_PYTHON_BINDINGS=ON -DAARE_TESTS=ON
+            cmake .. -DAARE_PYTHON_BINDINGS=ON -DAARE_TESTS=ON -DPython_FIND_VIRTUALENV=FIRST
             make -j 2 
 
       - name: C++ unit tests

.gitea/workflows/rh9-native.yml

@@ -8,7 +8,7 @@ permissions:
   contents: read
 
 jobs:
-  buildh:
+  build:
     runs-on: "ubuntu-latest"
     container:
       image: gitea.psi.ch/images/rhel9-developer-gitea-actions

.github/workflows/build_and_deploy_conda.yml

@@ -1,9 +1,9 @@
 name: Build pkgs and deploy if on main
 
 on:
-  push:
+  release:
-    branches:
+      types:
-     - main
+        - published
 
 jobs:
   build:
@@ -24,13 +24,13 @@ 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
-
+          conda-remove-defaults: "true"
-      - name: Prepare
-        run: conda install conda-build=24.9 conda-verify pytest anaconda-client
 
       - name: Enable upload
         run: conda config --set anaconda_upload yes

.github/workflows/build_conda.yml

@@ -24,13 +24,14 @@ 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
+          conda-remove-defaults: "true"
 
-      - name: Prepare
-        run: conda install conda-build=24.9 conda-verify pytest anaconda-client
   
       - name: Disable upload
         run: conda config --set anaconda_upload no

.github/workflows/build_docs.yml

@@ -2,7 +2,10 @@ name: Build the package using cmake then documentation
 
 on:
   workflow_dispatch:
-  push:
+  pull_request:
+  release:
+    types:
+      - published
     
 
 permissions:
@@ -40,7 +43,7 @@ jobs:
         run: |
             mkdir build 
             cd build 
-            cmake .. -DAARE_SYSTEM_LIBRARIES=ON -DAARE_DOCS=ON 
+            cmake .. -DAARE_SYSTEM_LIBRARIES=ON -DAARE_PYTHON_BINDINGS=ON -DAARE_DOCS=ON 
             make -j 2 
             make docs
 
@@ -55,7 +58,7 @@ jobs:
       url: ${{ steps.deployment.outputs.page_url }}
     runs-on: ubuntu-latest
     needs: build
-    if: github.ref == 'refs/heads/main'
+    if: (github.event_name == 'release' && github.event.action == 'published') || (github.event_name == 'workflow_dispatch' )
     steps:
       - name: Deploy to GitHub Pages
         id: deployment

.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

@@ -1,16 +1,29 @@
-cmake_minimum_required(VERSION 3.14)
+cmake_minimum_required(VERSION 3.15)
 
 project(aare 
-    VERSION 1.0.0
     DESCRIPTION "Data processing library for PSI detectors"
     HOMEPAGE_URL "https://github.com/slsdetectorgroup/aare"
     LANGUAGES C CXX 
 )
 
+# Read VERSION file into project version
+set(VERSION_FILE "${CMAKE_CURRENT_SOURCE_DIR}/VERSION")
+file(READ "${VERSION_FILE}" VERSION_CONTENT)
+string(STRIP "${VERSION_CONTENT}" PROJECT_VERSION_STRING)
+set(PROJECT_VERSION ${PROJECT_VERSION_STRING})
+
 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()
@@ -64,19 +77,12 @@ if(AARE_SYSTEM_LIBRARIES)
     # on conda-forge
 endif()
 
-if(AARE_VERBOSE)
-    add_compile_definitions(AARE_VERBOSE)
-endif()
-
-if(AARE_CUSTOM_ASSERT)
-    add_compile_definitions(AARE_CUSTOM_ASSERT)
-endif()
-
 if(AARE_BENCHMARKS)
     add_subdirectory(benchmarks)
 endif()
 
 
+
 set(CMAKE_EXPORT_COMPILE_COMMANDS ON)
 
 if(AARE_FETCH_LMFIT)
@@ -318,13 +324,10 @@ if(AARE_ASAN)
     )
 endif()
 
-
-
-
-
 if(AARE_TESTS)
     enable_testing()
     add_subdirectory(tests)
+    target_compile_definitions(tests PRIVATE AARE_TESTS)
 endif()
 
 ###------------------------------------------------------------------------------MAIN LIBRARY
@@ -347,8 +350,9 @@ set(PUBLICHEADERS
     include/aare/FilePtr.hpp
     include/aare/Frame.hpp
     include/aare/GainMap.hpp
-    include/aare/geo_helpers.hpp
+    include/aare/DetectorGeometry.hpp
     include/aare/JungfrauDataFile.hpp
+    include/aare/logger.hpp
     include/aare/NDArray.hpp
     include/aare/NDView.hpp
     include/aare/NumpyFile.hpp
@@ -372,7 +376,7 @@ set(SourceFiles
     ${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/DetectorGeometry.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/JungfrauDataFile.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/NumpyFile.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/src/NumpyHelpers.cpp
@@ -382,6 +386,7 @@ set(SourceFiles
     ${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/ifstream_helpers.cpp
 )
 
 add_library(aare_core STATIC ${SourceFiles})
@@ -390,6 +395,9 @@ target_include_directories(aare_core PUBLIC
     "$<INSTALL_INTERFACE:${CMAKE_INSTALL_INCLUDEDIR}>" 
 )
 
+set(THREADS_PREFER_PTHREAD_FLAG ON)
+find_package(Threads REQUIRED)
+
 target_link_libraries(
     aare_core 
     PUBLIC 
@@ -398,10 +406,25 @@ target_link_libraries(
     ${STD_FS_LIB} # from helpers.cmake
     PRIVATE 
     aare_compiler_flags 
+    Threads::Threads
     $<BUILD_INTERFACE:lmfit>
 
 )
 
+if(AARE_TESTS)
+    target_compile_definitions(aare_core PRIVATE AARE_TESTS)
+endif()
+if(AARE_VERBOSE)
+    target_compile_definitions(aare_core PUBLIC AARE_VERBOSE)
+    target_compile_definitions(aare_core PUBLIC AARE_LOG_LEVEL=aare::logDEBUG5)
+else()
+    target_compile_definitions(aare_core PUBLIC AARE_LOG_LEVEL=aare::logERROR)
+endif()
+
+if(AARE_CUSTOM_ASSERT)
+    target_compile_definitions(aare_core PUBLIC AARE_CUSTOM_ASSERT)
+endif()
+
 set_target_properties(aare_core PROPERTIES
     ARCHIVE_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}
     PUBLIC_HEADER "${PUBLICHEADERS}"
@@ -415,9 +438,10 @@ 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
+            ${CMAKE_CURRENT_SOURCE_DIR}/src/DetectorGeometry.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/RawMasterFile.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/NDArray.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/NDView.test.cpp
@@ -426,23 +450,19 @@ if(AARE_TESTS)
             ${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
+            ${CMAKE_CURRENT_SOURCE_DIR}/src/RawSubFile.test.cpp
             ${CMAKE_CURRENT_SOURCE_DIR}/src/utils/task.test.cpp
 
         )
     target_sources(tests PRIVATE ${TestSources} )
 endif()
 
-
-
-###------------------------------------------------------------------------------------------
-###------------------------------------------------------------------------------------------
-
-
 if(AARE_MASTER_PROJECT)
     install(TARGETS aare_core aare_compiler_flags 
         EXPORT "${TARGETS_EXPORT_NAME}"

RELEASE.md

@@ -0,0 +1,24 @@
+# Release notes
+
+
+### head
+
+Features:
+
+- Cluster finder now works with 5x5, 7x7 and 9x9 clusters
+- Added ClusterVector::empty() member
+
+
+### 2025.05.22
+
+Features:
+
+- Added scurve fitting
+
+Bugfixes:
+
+- Fixed crash when opening raw files with large number of data files
+- Fixed reading RawFiles with ROI fully excluding some sub files. 
+
+
+

VERSION

@@ -0,0 +1 @@
+2025.5.22

benchmarks/calculateeta_benchmark.cpp

@@ -41,8 +41,8 @@ BENCHMARK_F(ClusterFixture, Calculate2x2Eta)(benchmark::State &st) {
 }
 
 // almost takes double the time
-BENCHMARK_F(ClusterFixture,
+BENCHMARK_F(ClusterFixture, CalculateGeneralEtaFor2x2Cluster)
-            CalculateGeneralEtaFor2x2Cluster)(benchmark::State &st) {
+(benchmark::State &st) {
     for (auto _ : st) {
         // This code gets timed
         Eta2 eta = calculate_eta2<int, 2, 2>(cluster_2x2);
@@ -59,8 +59,8 @@ BENCHMARK_F(ClusterFixture, Calculate3x3Eta)(benchmark::State &st) {
 }
 
 // almost takes double the time
-BENCHMARK_F(ClusterFixture,
+BENCHMARK_F(ClusterFixture, CalculateGeneralEtaFor3x3Cluster)
-            CalculateGeneralEtaFor3x3Cluster)(benchmark::State &st) {
+(benchmark::State &st) {
     for (auto _ : st) {
         // This code gets timed
         Eta2 eta = calculate_eta2<int, 3, 3>(cluster_3x3);

benchmarks/ndarray_benchmark.cpp

@@ -1,19 +1,18 @@
-#include <benchmark/benchmark.h>
 #include "aare/NDArray.hpp"
-
+#include <benchmark/benchmark.h>
 
 using aare::NDArray;
 
 constexpr ssize_t size = 1024;
 class TwoArrays : public benchmark::Fixture {
-public:
+  public:
-    NDArray<int,2> a{{size,size},0};
+    NDArray<int, 2> a{{size, size}, 0};
-    NDArray<int,2> b{{size,size},0};
+    NDArray<int, 2> b{{size, size}, 0};
-  void SetUp(::benchmark::State& state) {
+    void SetUp(::benchmark::State &state) {
-    for(uint32_t i = 0; i < size; i++){
+        for (uint32_t i = 0; i < size; i++) {
-      for(uint32_t j = 0; j < size; j++){
+            for (uint32_t j = 0; j < size; j++) {
-        a(i, j)= i*j+1;
+                a(i, j) = i * j + 1;
-        b(i, j)= i*j+1;
+                b(i, j) = i * j + 1;
             }
         }
     }
@@ -22,20 +21,17 @@ public:
     // }
 };
 
-
+BENCHMARK_F(TwoArrays, AddWithOperator)(benchmark::State &st) {
-
-
-BENCHMARK_F(TwoArrays, AddWithOperator)(benchmark::State& st) {
     for (auto _ : st) {
         // This code gets timed
-    NDArray<int,2> res = a+b;
+        NDArray<int, 2> res = a + b;
         benchmark::DoNotOptimize(res);
     }
 }
-BENCHMARK_F(TwoArrays, AddWithIndex)(benchmark::State& st) {
+BENCHMARK_F(TwoArrays, AddWithIndex)(benchmark::State &st) {
     for (auto _ : st) {
         // This code gets timed
-    NDArray<int,2> res(a.shape());
+        NDArray<int, 2> res(a.shape());
         for (uint32_t i = 0; i < a.size(); i++) {
             res(i) = a(i) + b(i);
         }
@@ -43,17 +39,17 @@ BENCHMARK_F(TwoArrays, AddWithIndex)(benchmark::State& st) {
     }
 }
 
-BENCHMARK_F(TwoArrays, SubtractWithOperator)(benchmark::State& st) {
+BENCHMARK_F(TwoArrays, SubtractWithOperator)(benchmark::State &st) {
     for (auto _ : st) {
         // This code gets timed
-    NDArray<int,2> res = a-b;
+        NDArray<int, 2> res = a - b;
         benchmark::DoNotOptimize(res);
     }
 }
-BENCHMARK_F(TwoArrays, SubtractWithIndex)(benchmark::State& st) {
+BENCHMARK_F(TwoArrays, SubtractWithIndex)(benchmark::State &st) {
     for (auto _ : st) {
         // This code gets timed
-    NDArray<int,2> res(a.shape());
+        NDArray<int, 2> res(a.shape());
         for (uint32_t i = 0; i < a.size(); i++) {
             res(i) = a(i) - b(i);
         }
@@ -61,17 +57,17 @@ BENCHMARK_F(TwoArrays, SubtractWithIndex)(benchmark::State& st) {
     }
 }
 
-BENCHMARK_F(TwoArrays, MultiplyWithOperator)(benchmark::State& st) {
+BENCHMARK_F(TwoArrays, MultiplyWithOperator)(benchmark::State &st) {
     for (auto _ : st) {
         // This code gets timed
-    NDArray<int,2> res = a*b;
+        NDArray<int, 2> res = a * b;
         benchmark::DoNotOptimize(res);
     }
 }
-BENCHMARK_F(TwoArrays, MultiplyWithIndex)(benchmark::State& st) {
+BENCHMARK_F(TwoArrays, MultiplyWithIndex)(benchmark::State &st) {
     for (auto _ : st) {
         // This code gets timed
-    NDArray<int,2> res(a.shape());
+        NDArray<int, 2> res(a.shape());
         for (uint32_t i = 0; i < a.size(); i++) {
             res(i) = a(i) * b(i);
         }
@@ -79,17 +75,17 @@ BENCHMARK_F(TwoArrays, MultiplyWithIndex)(benchmark::State& st) {
     }
 }
 
-BENCHMARK_F(TwoArrays, DivideWithOperator)(benchmark::State& st) {
+BENCHMARK_F(TwoArrays, DivideWithOperator)(benchmark::State &st) {
     for (auto _ : st) {
         // This code gets timed
-    NDArray<int,2> res = a/b;
+        NDArray<int, 2> res = a / b;
         benchmark::DoNotOptimize(res);
     }
 }
-BENCHMARK_F(TwoArrays, DivideWithIndex)(benchmark::State& st) {
+BENCHMARK_F(TwoArrays, DivideWithIndex)(benchmark::State &st) {
     for (auto _ : st) {
         // This code gets timed
-    NDArray<int,2> res(a.shape());
+        NDArray<int, 2> res(a.shape());
         for (uint32_t i = 0; i < a.size(); i++) {
             res(i) = a(i) / b(i);
         }
@@ -97,17 +93,17 @@ BENCHMARK_F(TwoArrays, DivideWithIndex)(benchmark::State& st) {
     }
 }
 
-BENCHMARK_F(TwoArrays, FourAddWithOperator)(benchmark::State& st) {
+BENCHMARK_F(TwoArrays, FourAddWithOperator)(benchmark::State &st) {
     for (auto _ : st) {
         // This code gets timed
-    NDArray<int,2> res = a+b+a+b;
+        NDArray<int, 2> res = a + b + a + b;
         benchmark::DoNotOptimize(res);
     }
 }
-BENCHMARK_F(TwoArrays, FourAddWithIndex)(benchmark::State& st) {
+BENCHMARK_F(TwoArrays, FourAddWithIndex)(benchmark::State &st) {
     for (auto _ : st) {
         // This code gets timed
-    NDArray<int,2> res(a.shape());
+        NDArray<int, 2> res(a.shape());
         for (uint32_t i = 0; i < a.size(); i++) {
             res(i) = a(i) + b(i) + a(i) + b(i);
         }
@@ -115,17 +111,17 @@ BENCHMARK_F(TwoArrays, FourAddWithIndex)(benchmark::State& st) {
     }
 }
 
-BENCHMARK_F(TwoArrays, MultiplyAddDivideWithOperator)(benchmark::State& st) {
+BENCHMARK_F(TwoArrays, MultiplyAddDivideWithOperator)(benchmark::State &st) {
     for (auto _ : st) {
         // This code gets timed
-    NDArray<int,2> res = a*a+b/a;
+        NDArray<int, 2> res = a * a + b / a;
         benchmark::DoNotOptimize(res);
     }
 }
-BENCHMARK_F(TwoArrays, MultiplyAddDivideWithIndex)(benchmark::State& st) {
+BENCHMARK_F(TwoArrays, MultiplyAddDivideWithIndex)(benchmark::State &st) {
     for (auto _ : st) {
         // This code gets timed
-    NDArray<int,2> res(a.shape());
+        NDArray<int, 2> res(a.shape());
         for (uint32_t i = 0; i < a.size(); i++) {
             res(i) = a(i) * a(i) + b(i) / a(i);
         }

conda-recipe/conda_build_config.yaml

@@ -1,28 +1,5 @@
 python:
     - 3.11
-    - 3.11
-    - 3.11
-    - 3.12
-    - 3.12
     - 3.12
     - 3.13
 
-
-    
-numpy: 
-    - 1.26
-    - 2.0
-    - 2.1
-    - 1.26
-    - 2.0
-    - 2.1
-    - 2.1
-
-
-zip_keys:
-    - python
-    - numpy
-
-pin_run_as_build:
-  numpy: x.x
-  python: x.x

conda-recipe/meta.yaml

@@ -1,10 +1,10 @@
+source:
+  path: ../
+
+{% set version = load_file_regex(load_file = 'VERSION', regex_pattern = '(\d+(?:\.\d+)*(?:[\+\w\.]+))').group(1) %}
 package:
   name: aare
-  version: 2025.4.1 #TODO! how to not duplicate this?
+  version: {{version}}
-
-
-
-
 
 source:
   path: ..
@@ -12,44 +12,39 @@ source:
 build:
   number: 0
   script:
-    - unset CMAKE_GENERATOR && {{ PYTHON }} -m pip install . -vv  # [not win]
+    - unset CMAKE_GENERATOR && {{ PYTHON }} -m pip install . -vv 
-    - {{ PYTHON }} -m pip install . -vv  # [win]
 
 requirements:
   build:
-    - python {{python}}
-    - numpy {{ numpy }}
     - {{ compiler('cxx') }}
-
-
-  host:
     - cmake
     - ninja
-    - python {{python}}
+
-    - numpy {{ numpy }}
+  host:
+    - python
     - pip
+    - numpy=2.1
     - scikit-build-core
     - pybind11 >=2.13.0
-    - fmt
+    - matplotlib # needed in host to solve the environment for run
-    - zeromq
-    - nlohmann_json
-    - catch2
 
   run:
-    - python {{python}}
+    - python
-    - numpy {{ numpy }}
+    - {{ pin_compatible('numpy') }}
+    - matplotlib
+    
 
 
 test:
   imports:
     - aare
-  # requires:
+  requires:
-  #   - pytest
+    - pytest
-  # source_files:
+    - boost-histogram 
-  #   - tests
+  source_files:
-  # commands:
+    - python/tests
-  #   - pytest tests
+  commands:
+    - python -m pytest python/tests
 
 about:
-  summary: An example project built with pybind11 and scikit-build.
+  summary: Data analysis library for hybrid pixel detectors from PSI 
-  # license_file: LICENSE

docs/CMakeLists.txt

@@ -45,12 +45,3 @@ add_custom_target(
     COMMENT "Generating documentation with Sphinx"
 )
 
-add_custom_target(
-    rst
-    COMMAND ${SPHINX_EXECUTABLE} -a -b html
-    -Dbreathe_projects.aare=${CMAKE_CURRENT_BINARY_DIR}/xml
-    -c "${SPHINX_BUILD}"
-    ${SPHINX_BUILD}/src 
-    ${SPHINX_BUILD}/html
-    COMMENT "Generating documentation with Sphinx"
-)

docs/Doxyfile.in

@@ -886,7 +886,7 @@ EXCLUDE_SYMLINKS       = NO
 # Note that the wildcards are matched against the file with absolute path, so to
 # exclude all test directories for example use the pattern */test/*
 
-EXCLUDE_PATTERNS       = */docs/* */tests/* */python/* */manual */slsDetectorServers/* */libs/* */integrationTests *README* */slsDetectorGui/* */ctbGui/* */slsDetectorCalibration/* *TobiSchluter*
+EXCLUDE_PATTERNS       = *build* */docs/* */tests/* *.test.cpp* */python/* */manual */slsDetectorServers/* */libs/* */integrationTests *README* *_deps* *TobiSchluter*
 
 # The EXCLUDE_SYMBOLS tag can be used to specify one or more symbol names
 # (namespaces, classes, functions, etc.) that should be excluded from the

docs/src/ClusterFile.rst

@@ -5,3 +5,4 @@ ClusterFile
    :members:
    :undoc-members:
    :private-members:
+

docs/src/ClusterVector.rst

@@ -4,3 +4,12 @@ ClusterVector
 .. doxygenclass:: aare::ClusterVector
    :members:
    :undoc-members:
+   :private-members:
+
+
+
+.. doxygenclass:: aare::ClusterVector< Cluster< T, ClusterSizeX, ClusterSizeY, CoordType > > 
+   :members:
+   :undoc-members:
+   :private-members:
+   

docs/src/Philosophy.rst

@@ -0,0 +1,47 @@
+****************
+Philosophy
+****************
+
+
+Fast code with a simple interface
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Aare should be fast and efficient, but also easy to use. We strive to keep a simple interface that feels intuitive. 
+Internally we use C++ for performance and the ability to integrate the library in other programs, but we see most 
+users using the Python interface. 
+
+Live at head
+~~~~~~~~~~~~~~~~~~
+
+As a user of the library you should be able to, and is expected to, use the latest version. Bug fixes will rarely be backported
+to older releases. By upgrading frequently you will benefit from the latest features and minimize the effort to maintain your scripts/code 
+by doing several small upgrades instead of one big upgrade.
+
+API
+~~~~~~~~~~~~~~~~~~
+
+We aim to keep the API stable and only break it for good reasons. But specially now in the early stages of development
+the API will change. On those occasions it will be clearly stated in the release notes. However, the norm should be a 
+backward compatible API.
+
+Documentation
+~~~~~~~~~~~~~~~~~~
+
+Being a library it is important to have a well documented API. We use Doxygen to generate the C++ documentation
+and Sphinx for the Python part. Breathe is used to integrate the two into one Sphinx html site. The documentation is built
+automatically on release by the CI and published to GitHub pages. In addition to the generated API documentation,
+certain classes might need more descriptions of the usage. This is then placed in the .rst files in the docs/src directory.
+
+.. attention::
+
+    The code should be well documented, but using descriptive names is more important. In the same spirit
+    if a function is called `getNumberOfFrames()` you don't need to write a comment saying that it gets the 
+    number of frames.
+
+
+Dependencies
+~~~~~~~~~~~~~~~~~~
+
+Deployment in the scientific community is often tricky. Either due to old OS versions or the lack of package managers. 
+We strive to keep the dependencies to a minimum and will vendor some libraries to simplify deployment even though it comes
+at a cost of build time.

docs/src/Requirements.rst

@@ -2,18 +2,21 @@ Requirements
 ==============================================
 
 - C++17 compiler (gcc 8/clang 7)
-- CMake 3.14+
+- CMake 3.15+
 
 **Internally used libraries**
 
 .. note ::
 
-    These can also be picked up from the system/conda environment by specifying:
+    To save compile time some of the dependencies can also be picked up from the system/conda environment by specifying:
     -DAARE_SYSTEM_LIBRARIES=ON during the cmake configuration.
 
-- pybind11
+To simplify deployment we build and statically link a few libraries.
+
 - fmt
+- lmfit - https://jugit.fz-juelich.de/mlz/lmfit
 - nlohmann_json
+- pybind11
 - ZeroMQ
 
 **Extra dependencies for building documentation**

docs/src/Tests.rst

@@ -2,7 +2,7 @@
 Tests
 ****************
 
-We test the code both from the C++ and Python API. By default only tests that does not require image data is run. 
+We test the code both from C++ and Python. By default only tests that does not require additional data are run.
 
 C++
 ~~~~~~~~~~~~~~~~~~
@@ -15,7 +15,7 @@ C++
     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
+    ./run_test [.with-data] #or using ctest, [.with-data] is the option to include tests needing data
 
 
 
@@ -25,7 +25,7 @@ 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
+    python -m pytest python/tests --with-data # passing --with-data will run the tests needing data
 
 
 
@@ -35,7 +35,7 @@ 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
+    AARE_TEST_DATA to the path of the test data directory. Then pass either [.with-data] 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.

docs/src/Workflow.rst

@@ -0,0 +1,86 @@
+****************
+Workflow
+****************
+
+This page describes how we develop aare. 
+
+GitHub centric
+~~~~~~~~~~~~~~~~~~
+
+We use GitHub for all development. Issues and pull requests provide a platform for collaboration as well
+as a record of the development process. Even if we discuss things in person, we record the outcome in an issue.
+If a particular implementation is chosen over another, the reason should be recorded in the pull request.
+
+
+Branches
+~~~~~~~~~~~~~~~~~~
+
+We aim for an as lightweight branching strategy as possible. Short-lived feature branches are merged back into main. 
+The main branch is expected to always be in a releasable state. A release is simply a tag on main which provides a
+reference and triggers the CI to build the release artifacts (conda, pypi etc.). For large features consider merging
+smaller chunks into main as they are completed, rather than waiting for the entire feature to be finished. Worst case 
+make sure your feature branch merges with main regularly to avoid large merge conflicts later on.
+
+.. note::
+
+    The main branch is expected to always work. Feel free to pull from main instead of sticking to a 
+    release
+
+
+Releases
+~~~~~~~~~~~~~~~~~~
+
+Release early, release often. As soon as "enough" new features have been implemented, a release is created.
+A release should not be a big thing, rather a routine part of development that does not require any special person or
+unfamiliar steps.
+
+
+
+Checklists for deployment
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+**Feature:**
+
+#. Create a new issue for the feature (label feature)
+#. Create a new branch from main.
+#. Implement the feature including test and documentation
+#. Add the feature to RELEASE.md under head
+#. Create a pull request linked to the issue
+#. Code is reviewed by at least one other person
+#. Once approved, the branch is merged into main
+
+
+**BugFix:**
+
+Essentially the same as for a feature, if possible start with
+a failing test that demonstrates the bug.
+
+#. Create a new issue for the bug (label bug)
+#. Create a new branch from main.
+#. **Write a test that fails for the bug**
+#. Implement the fix
+#. **Run the test to ensure it passes**
+#. Add the bugfix to RELEASE.md under head
+#. Create a pull request linked to the issue.
+#. Code is reviewed by at least one other person
+#. Once approved, the branch is merged into main
+
+**Release:**
+
+#. Once "enough" new features have been implemented, a release is created
+#. Update RELEASE.md with the tag of the release and verify that it is complete
+#. Create the release in GitHub describing the new features and bug fixes
+#. CI makes magic
+
+
+**Update documentation only:**
+
+.. attention::
+
+    It's possible to update the documentation without changing the code, but take
+    care since the docs will reflect the code in main and not the latest release.
+
+#. Create a PR to main with the documentation changes
+#. Create a pull request linked to the issue.
+#. Code is reviewed by at least one other person
+#. Once merged you can manually trigger the CI workflow for documentation

docs/src/index.rst

@@ -63,4 +63,6 @@ AARE
     :caption: Developer
     :maxdepth: 3
     
+    Philosophy
+    Workflow
     Tests

docs/src/pyClusterFile.rst

@@ -2,9 +2,24 @@
 ClusterFile
 ============
 
+
+The :class:`ClusterFile` class is the main interface to read and write clusters in aare. Unfortunately the
+old file format does not include metadata like the cluster size and the data type. This means that the
+user has to know this information from other sources. Specifying the wrong cluster size or data type
+will lead to garbage data being read. 
+
 .. py:currentmodule:: aare
 
 .. autoclass:: ClusterFile
+    :members:
+    :undoc-members:
+    :inherited-members:
+
+
+Below is the API of the ClusterFile_Cluster3x3i but all variants share the same API.
+
+.. autoclass:: aare._aare.ClusterFile_Cluster3x3i
+    :special-members: __init__
     :members:
     :undoc-members:
     :show-inheritance:

docs/src/pyClusterVector.rst

@@ -2,8 +2,10 @@ ClusterVector
 ================
 
 The ClusterVector, holds clusters from the ClusterFinder. Since it is templated
-in C++  we use a suffix indicating the data type in python. The suffix is
+in C++  we use a suffix indicating the type of cluster it holds. The suffix follows
-``_i`` for integer, ``_f`` for float, and ``_d`` for double.
+the same pattern as for ClusterFile i.e. ``ClusterVector_Cluster3x3i``
+for a vector holding 3x3 integer clusters.
+
 
 At the moment the functionality from python is limited and it is not supported
 to push_back clusters to the vector. The intended use case is to pass it to 
@@ -26,7 +28,8 @@ C++ functions that support the ClusterVector or to view it as a numpy array.
 
 .. py:currentmodule:: aare
 
-.. autoclass:: ClusterVector_i
+.. autoclass::  aare._aare.ClusterVector_Cluster3x3i
+    :special-members: __init__
     :members:
     :undoc-members:
     :show-inheritance:

etc/dev-env.yml

@@ -3,13 +3,14 @@ channels:
   - conda-forge
 dependencies:
   - anaconda-client
+  - conda-build
   - doxygen 
-  - sphinx=7.1.2 
+  - sphinx
   - breathe 
-  - pybind11 
   - sphinx_rtd_theme 
   - furo 
-  - nlohmann_json 
   - zeromq 
-  - fmt 
+  - pybind11
   - numpy
+  - matplotlib
+

include/aare/ArrayExpr.hpp

@@ -1,22 +1,25 @@
 #pragma once
-#include <cstdint> //int64_t
+#include "aare/defs.hpp"
-#include <cstddef> //size_t
 #include <array>
-
 #include <cassert>
+#include <cstddef>
+#include <cstdint>
+
 namespace aare {
 
-template <typename E, int64_t Ndim> class ArrayExpr {
+template <typename E, ssize_t Ndim> class ArrayExpr {
   public:
     static constexpr bool is_leaf = false;
 
     auto operator[](size_t i) const { return static_cast<E const &>(*this)[i]; }
     auto operator()(size_t i) const { return static_cast<E const &>(*this)[i]; }
     auto size() const { return static_cast<E const &>(*this).size(); }
-    std::array<int64_t, Ndim> shape() const { return static_cast<E const &>(*this).shape(); }
+    std::array<ssize_t, Ndim> shape() const {
+        return static_cast<E const &>(*this).shape();
+    }
 };
 
-template <typename A, typename B, int64_t Ndim>
+template <typename A, typename B, ssize_t Ndim>
 class ArrayAdd : public ArrayExpr<ArrayAdd<A, B, Ndim>, Ndim> {
     const A &arr1_;
     const B &arr2_;
@@ -27,10 +30,10 @@ class ArrayAdd : public ArrayExpr<ArrayAdd<A, B, Ndim>, Ndim> {
     }
     auto operator[](int i) const { return arr1_[i] + arr2_[i]; }
     size_t size() const { return arr1_.size(); }
-    std::array<int64_t, Ndim> shape() const { return arr1_.shape(); }
+    std::array<ssize_t, Ndim> shape() const { return arr1_.shape(); }
 };
 
-template <typename A, typename B, int64_t Ndim>
+template <typename A, typename B, ssize_t Ndim>
 class ArraySub : public ArrayExpr<ArraySub<A, B, Ndim>, Ndim> {
     const A &arr1_;
     const B &arr2_;
@@ -41,11 +44,11 @@ class ArraySub : public ArrayExpr<ArraySub<A, B, Ndim>, Ndim> {
     }
     auto operator[](int i) const { return arr1_[i] - arr2_[i]; }
     size_t size() const { return arr1_.size(); }
-    std::array<int64_t, Ndim> shape() const { return arr1_.shape(); }
+    std::array<ssize_t, Ndim> shape() const { return arr1_.shape(); }
 };
 
-template <typename A, typename B, int64_t Ndim>
+template <typename A, typename B, ssize_t Ndim>
-class ArrayMul : public ArrayExpr<ArrayMul<A, B, Ndim>,Ndim> {
+class ArrayMul : public ArrayExpr<ArrayMul<A, B, Ndim>, Ndim> {
     const A &arr1_;
     const B &arr2_;
 
@@ -55,10 +58,10 @@ class ArrayMul : public ArrayExpr<ArrayMul<A, B, Ndim>,Ndim> {
     }
     auto operator[](int i) const { return arr1_[i] * arr2_[i]; }
     size_t size() const { return arr1_.size(); }
-    std::array<int64_t, Ndim> shape() const { return arr1_.shape(); }
+    std::array<ssize_t, Ndim> shape() const { return arr1_.shape(); }
 };
 
-template <typename A, typename B, int64_t Ndim>
+template <typename A, typename B, ssize_t Ndim>
 class ArrayDiv : public ArrayExpr<ArrayDiv<A, B, Ndim>, Ndim> {
     const A &arr1_;
     const B &arr2_;
@@ -69,31 +72,27 @@ class ArrayDiv : public ArrayExpr<ArrayDiv<A, B, Ndim>, Ndim> {
     }
     auto operator[](int i) const { return arr1_[i] / arr2_[i]; }
     size_t size() const { return arr1_.size(); }
-    std::array<int64_t, Ndim> shape() const { return arr1_.shape(); }
+    std::array<ssize_t, Ndim> shape() const { return arr1_.shape(); }
 };
 
-
+template <typename A, typename B, ssize_t Ndim>
-
-template <typename A, typename B, int64_t Ndim>
 auto operator+(const ArrayExpr<A, Ndim> &arr1, const ArrayExpr<B, Ndim> &arr2) {
     return ArrayAdd<ArrayExpr<A, Ndim>, ArrayExpr<B, Ndim>, Ndim>(arr1, arr2);
 }
 
-template <typename A, typename B, int64_t Ndim>
+template <typename A, typename B, ssize_t Ndim>
-auto operator-(const ArrayExpr<A,Ndim> &arr1, const ArrayExpr<B, Ndim> &arr2) {
+auto operator-(const ArrayExpr<A, Ndim> &arr1, const ArrayExpr<B, Ndim> &arr2) {
     return ArraySub<ArrayExpr<A, Ndim>, ArrayExpr<B, Ndim>, Ndim>(arr1, arr2);
 }
 
-template <typename A, typename B, int64_t Ndim>
+template <typename A, typename B, ssize_t Ndim>
 auto operator*(const ArrayExpr<A, Ndim> &arr1, const ArrayExpr<B, Ndim> &arr2) {
     return ArrayMul<ArrayExpr<A, Ndim>, ArrayExpr<B, Ndim>, Ndim>(arr1, arr2);
 }
 
-template <typename A, typename B, int64_t Ndim>
+template <typename A, typename B, ssize_t Ndim>
 auto operator/(const ArrayExpr<A, Ndim> &arr1, const ArrayExpr<B, Ndim> &arr2) {
     return ArrayDiv<ArrayExpr<A, Ndim>, ArrayExpr<B, Ndim>, Ndim>(arr1, arr2);
 }
 
-
-
 } // namespace aare

include/aare/CalculateEta.hpp

@@ -6,14 +6,14 @@
 
 namespace aare {
 
-typedef enum {
+enum class corner : int {
     cBottomLeft = 0,
     cBottomRight = 1,
     cTopLeft = 2,
     cTopRight = 3
-} corner;
+};
 
-typedef enum {
+enum class pixel : int {
     pBottomLeft = 0,
     pBottom = 1,
     pBottomRight = 2,
@@ -23,7 +23,7 @@ typedef enum {
     pTopLeft = 6,
     pTop = 7,
     pTopRight = 8
-} pixel;
+};
 
 template <typename T> struct Eta2 {
     double x;
@@ -41,7 +41,7 @@ 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.at(i));
+        auto e = calculate_eta2(clusters[i]);
         eta2(i, 0) = e.x;
         eta2(i, 1) = e.y;
     }
@@ -64,31 +64,79 @@ calculate_eta2(const Cluster<T, ClusterSizeX, ClusterSizeY, CoordType> &cl) {
     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);
 
-    if ((cl.data[index_bottom_left_max_2x2_subcluster] +
+    // check that cluster center is in max subcluster
-         cl.data[index_bottom_left_max_2x2_subcluster + 1]) != 0)
+    if (cluster_center_index != index_bottom_left_max_2x2_subcluster &&
-        eta.x = static_cast<double>(
+        cluster_center_index != index_bottom_left_max_2x2_subcluster + 1 &&
-                    cl.data[index_bottom_left_max_2x2_subcluster + 1]) /
+        cluster_center_index !=
-                static_cast<double>(
+            index_bottom_left_max_2x2_subcluster + ClusterSizeX &&
-                    (cl.data[index_bottom_left_max_2x2_subcluster] +
+        cluster_center_index !=
-                     cl.data[index_bottom_left_max_2x2_subcluster + 1]));
+            index_bottom_left_max_2x2_subcluster + ClusterSizeX + 1)
+        throw std::runtime_error("Photon center is not in max 2x2_subcluster");
 
-    if ((cl.data[index_bottom_left_max_2x2_subcluster] +
+    if ((cluster_center_index - index_bottom_left_max_2x2_subcluster) %
-         cl.data[index_bottom_left_max_2x2_subcluster + ClusterSizeX]) != 0)
+            ClusterSizeX ==
-        eta.y =
+        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[index_bottom_left_max_2x2_subcluster + ClusterSizeX]) /
+                        (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[index_bottom_left_max_2x2_subcluster] +
+                        (cl.data[cluster_center_index] +
-                 cl.data[index_bottom_left_max_2x2_subcluster + ClusterSizeX]));
+                         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) {

include/aare/CircularFifo.hpp

@@ -17,7 +17,8 @@ template <class ItemType> class CircularFifo {
 
   public:
     CircularFifo() : CircularFifo(100){};
-    CircularFifo(uint32_t size) : fifo_size(size), free_slots(size + 1), filled_slots(size + 1) {
+    CircularFifo(uint32_t size)
+        : fifo_size(size), free_slots(size + 1), filled_slots(size + 1) {
 
         // TODO! how do we deal with alignment for writing? alignas???
         // Do we give the user a chance to provide memory locations?
@@ -55,7 +56,8 @@ template <class ItemType> class CircularFifo {
 
     bool try_pop_free(ItemType &v) { return free_slots.read(v); }
 
-    ItemType pop_value(std::chrono::nanoseconds wait, std::atomic<bool> &stopped) {
+    ItemType pop_value(std::chrono::nanoseconds wait,
+                       std::atomic<bool> &stopped) {
         ItemType v;
         while (!filled_slots.read(v) && !stopped) {
             std::this_thread::sleep_for(wait);

include/aare/Cluster.hpp

@@ -16,28 +16,43 @@
 
 namespace aare {
 
-template <typename T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
-          typename CoordType = int16_t>
-constexpr bool is_valid_cluster =
-    std::is_arithmetic_v<T> && std::is_integral_v<CoordType> &&
-    (ClusterSizeX > 0) && (ClusterSizeY > 0);
-
 // requires clause c++20 maybe update
 template <typename T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
-          typename CoordType = int16_t,
+          typename CoordType = int16_t>
-          typename Enable = std::enable_if_t<
-              is_valid_cluster<T, ClusterSizeX, ClusterSizeY, CoordType>>>
 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;
-    T data[ClusterSizeX * ClusterSizeY];
+    std::array<T, ClusterSizeX * ClusterSizeY> data;
 
-    T sum() const {
+    static constexpr uint8_t cluster_size_x = ClusterSizeX;
-        return std::accumulate(data, data + ClusterSizeX * ClusterSizeY, 0);
+    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);
 
@@ -56,42 +71,112 @@ struct Cluster {
                         sum_2x2_subcluster.begin();
             return std::make_pair(sum_2x2_subcluster[index], index);
         }
-};
-
-// Specialization for 2x2 clusters (only one sum exists)
-template <typename T> struct Cluster<T, 2, 2, int16_t> {
-    int16_t x;
-    int16_t y;
-    T data[4];
-
-    T sum() const { return std::accumulate(data, data + 4, 0); }
-
-    std::pair<T, int> max_sum_2x2() const {
-        return std::make_pair(data[0] + data[1] + data[2] + data[3],
-                              0); // Only one possible 2x2 sum
     }
 };
 
-// Specialization for 3x3 clusters
+template<typename T>
-template <typename T> struct Cluster<T, 3, 3, int16_t> {
+Cluster<T, 2, 2, uint16_t> reduce_3x3_to_2x2(const Cluster<T, 3, 3, uint16_t> &c) {
-    int16_t x;
+    Cluster<T, 2, 2, uint16_t> result;
-    int16_t y;
-    T data[9];
     
-    T sum() const { return std::accumulate(data, data + 9, 0); }
+    auto [s, i] = c.max_sum_2x2();
-
+    switch (i) {
-    std::pair<T, int> max_sum_2x2() const {
+        case 0:
-        std::array<T, 4> sum_2x2_subclusters;
+            result.x = c.x-1;
-        sum_2x2_subclusters[0] = data[0] + data[1] + data[3] + data[4];
+            result.y = c.y+1;
-        sum_2x2_subclusters[1] = data[1] + data[2] + data[4] + data[5];
+            result.data = {c.data[0], c.data[1], c.data[3], c.data[4]};
-        sum_2x2_subclusters[2] = data[3] + data[4] + data[6] + data[7];
+            break;
-        sum_2x2_subclusters[3] = data[4] + data[5] + data[7] + data[8];
+        case 1:
-        int index = std::max_element(sum_2x2_subclusters.begin(),
+            result.x = c.x;
-                                     sum_2x2_subclusters.end()) -
+            result.y = c.y + 1;
-                    sum_2x2_subclusters.begin();
+            result.data = {c.data[1], c.data[2], c.data[4], c.data[5]};
-        return std::make_pair(sum_2x2_subclusters[index], index);
+            break;
+        case 2:
+            result.x = c.x -1;
+            result.y = c.y;
+            result.data = {c.data[3], c.data[4], c.data[6], c.data[7]};
+            break;
+        case 3:
+            result.x = c.x;
+            result.y = c.y;
+            result.data = {c.data[4], c.data[5], c.data[7], c.data[8]};
+            break;
     }
-};
+
+    // do some stuff
+    return result;
+}
+
+template<typename T>
+Cluster<T, 3, 3, uint16_t> reduce_5x5_to_3x3(const Cluster<T, 5, 5, uint16_t> &c) {
+    Cluster<T, 3, 3, uint16_t> result;
+
+    // Reduce the 5x5 cluster to a 3x3 cluster by selecting the 3x3 block with the highest sum
+    std::array<T, 9> sum_3x3_subclusters;
+
+    //Write out the sums in the hope that the compiler can optimize this
+    sum_3x3_subclusters[0] = c.data[0] + c.data[1] + c.data[2] + c.data[5] + c.data[6] + c.data[7] + c.data[10] + c.data[11] + c.data[12];
+    sum_3x3_subclusters[1] = c.data[1] + c.data[2] + c.data[3] + c.data[6] + c.data[7] + c.data[8] + c.data[11] + c.data[12] + c.data[13];
+    sum_3x3_subclusters[2] = c.data[2] + c.data[3] + c.data[4] + c.data[7] + c.data[8] + c.data[9] + c.data[12] + c.data[13] + c.data[14];
+    sum_3x3_subclusters[3] = c.data[5] + c.data[6] + c.data[7] + c.data[10] + c.data[11] + c.data[12] + c.data[15] + c.data[16] + c.data[17];
+    sum_3x3_subclusters[4] = c.data[6] + c.data[7] + c.data[8] + c.data[11] + c.data[12] + c.data[13] + c.data[16] + c.data[17] + c.data[18];
+    sum_3x3_subclusters[5] = c.data[7] + c.data[8] + c.data[9] + c.data[12] + c.data[13] + c.data[14] + c.data[17] + c.data[18] + c.data[19];
+    sum_3x3_subclusters[6] = c.data[10] + c.data[11] + c.data[12] + c.data[15] + c.data[16] + c.data[17] + c.data[20] + c.data[21] + c.data[22];
+    sum_3x3_subclusters[7] = c.data[11] + c.data[12] + c.data[13] + c.data[16] + c.data[17] + c.data[18] + c.data[21] + c.data[22] + c.data[23];
+    sum_3x3_subclusters[8] = c.data[12] + c.data[13] + c.data[14] + c.data[17] + c.data[18] + c.data[19] + c.data[22] + c.data[23] + c.data[24];
+
+    auto index = std::max_element(sum_3x3_subclusters.begin(), sum_3x3_subclusters.end()) - sum_3x3_subclusters.begin();
+
+    switch (index) {
+        case 0:
+            result.x = c.x - 1;
+            result.y = c.y + 1;
+            result.data = {c.data[0], c.data[1], c.data[2], c.data[5], c.data[6], c.data[7], c.data[10], c.data[11], c.data[12]};
+            break;
+        case 1:
+            result.x = c.x;
+            result.y = c.y + 1;
+            result.data = {c.data[1], c.data[2], c.data[3], c.data[6], c.data[7], c.data[8], c.data[11], c.data[12], c.data[13]};
+            break;
+        case 2:
+            result.x = c.x + 1;
+            result.y = c.y + 1;
+            result.data = {c.data[2], c.data[3], c.data[4], c.data[7], c.data[8], c.data[9], c.data[12], c.data[13], c.data[14]};
+            break;
+        case 3:
+            result.x = c.x - 1;
+            result.y = c.y;
+            result.data = {c.data[5], c.data[6], c.data[7], c.data[10], c.data[11], c.data[12], c.data[15], c.data[16], c.data[17]};
+            break;
+        case 4:
+            result.x = c.x + 1;
+            result.y = c.y;
+            result.data = {c.data[6], c.data[7], c.data[8], c.data[11], c.data[12], c.data[13], c.data[16], c.data[17], c.data[18]};
+            break;
+        case 5:
+            result.x = c.x + 1;
+            result.y = c.y;
+            result.data = {c.data[7], c.data[8], c.data[9], c.data[12], c.data[13], c.data[14], c.data[17], c.data[18], c.data[19]};
+            break;
+        case 6:
+            result.x = c.x + 1;
+            result.y = c.y -1;
+            result.data = {c.data[10], c.data[11], c.data[12], c.data[15], c.data[16], c.data[17], c.data[20], c.data[21], c.data[22]};
+            break;
+        case 7:
+            result.x = c.x + 1;
+            result.y = c.y-1;
+            result.data = {c.data[11], c.data[12], c.data[13], c.data[16], c.data[17], c.data[18], c.data[21], c.data[22], c.data[23]};
+            break;
+        case 8:
+            result.x = c.x + 1;
+            result.y = c.y-1;
+            result.data = {c.data[12], c.data[13], c.data[14], c.data[17], c.data[18], c.data[19], c.data[22], c.data[23], c.data[24]};
+            break;
+    }
+
+    // do some stuff
+    return result;
+}
 
 // Type Traits for is_cluster_type
 template <typename T>
@@ -102,20 +187,4 @@ struct is_cluster<Cluster<T, X, Y, CoordType>> : std::true_type {}; // Cluster
 
 template <typename T> constexpr bool is_cluster_v = is_cluster<T>::value;
 
-template <typename ClusterType,
-          typename = std::enable_if_t<is_cluster_v<ClusterType>>>
-struct extract_template_arguments; // Forward declaration
-
-// helper struct to extract template argument
-template <typename T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
-          typename CoordType>
-struct extract_template_arguments<
-    Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>> {
-
-    using value_type = T;
-    static constexpr int cluster_size_x = ClusterSizeX;
-    static constexpr int cluster_size_y = ClusterSizeY;
-    using coordtype = CoordType;
-};
-
 } // namespace aare

include/aare/ClusterCollector.hpp

@@ -37,7 +37,11 @@ class ClusterCollector {
   public:
     ClusterCollector(ClusterFinderMT<ClusterType, uint16_t, double> *source) {
         m_source = source->sink();
-        m_thread = std::thread(&ClusterCollector::process, this);
+        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;

include/aare/ClusterFile.hpp

@@ -5,6 +5,8 @@
 #include "aare/GainMap.hpp"
 #include "aare/NDArray.hpp"
 #include "aare/defs.hpp"
+#include "aare/logger.hpp"
+
 #include <filesystem>
 #include <fstream>
 #include <optional>
@@ -12,7 +14,7 @@
 namespace aare {
 
 /*
-Binary cluster file. Expects data to be layed out as:
+Binary cluster file. Expects data to be laid out as:
 int32_t frame_number
 uint32_t number_of_clusters
 int16_t x, int16_t y, int32_t data[9] x number_of_clusters
@@ -46,8 +48,8 @@ class ClusterFile {
     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<GainMap> m_gain_map; /*Gain map to apply to the clusters, will
+    std::optional<InvertedGainMap> m_gain_map; /*Gain map to apply to the
-                                          be applied if set*/
+                                          clusters, will be applied if set*/
 
   public:
     /**
@@ -60,80 +62,8 @@ 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");
+                const std::string &mode = "r")
 
-    ~ClusterFile();
-
-    /**
-     * @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<ClusterType> read_clusters(size_t 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
-     */
-    ClusterVector<ClusterType> read_frame();
-
-    void write_frame(const ClusterVector<ClusterType> &clusters);
-
-    /**
-     * @brief Return the chunk size
-     */
-    size_t chunk_size() const { return m_chunk_size; }
-
-    /**
-     * @brief Set the region of interest to use when reading clusters. If set
-     * only clusters within the ROI will be read.
-     */
-    void set_roi(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);
-
-    /**
-     * @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.
-     */
-    void set_gain_map(const NDView<double, 2> gain_map);
-
-    void set_gain_map(const GainMap &gain_map);
-
-    void set_gain_map(const GainMap &&gain_map);
-
-    /**
-     * @brief Close the file. If not closed the file will be closed in the
-     * destructor
-     */
-    void close();
-
-    /** @brief Open the file in specific mode
-     *
-     */
-    void open(const std::string &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();
-};
-
-template <typename ClusterType, typename Enable>
-ClusterFile<ClusterType, Enable>::ClusterFile(
-    const std::filesystem::path &fname, size_t chunk_size,
-    const std::string &mode)
         : m_filename(fname.string()), m_chunk_size(chunk_size), m_mode(mode) {
 
         if (mode == "r") {
@@ -157,23 +87,112 @@ ClusterFile<ClusterType, Enable>::ClusterFile(
         } else {
             throw std::runtime_error("Unsupported mode: " + mode);
         }
-}
+    }
 
-template <typename ClusterType, typename Enable>
+    ~ClusterFile() { close(); }
-ClusterFile<ClusterType, Enable>::~ClusterFile() {
-    close();
-}
 
-template <typename ClusterType, typename Enable>
+    /**
-void ClusterFile<ClusterType, Enable>::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
+     */
+    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
+     */
+    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");
+        }
+
+        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 Set the region of interest to use when reading
+     * clusters. If set only clusters within the ROI will be
+     * read.
+     */
+    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;
         }
-}
+    }
 
-template <typename ClusterType, typename Enable>
+    /** @brief Open the file in specific mode
-void ClusterFile<ClusterType, Enable>::open(const std::string &mode) {
+     *
+     */
+    void open(const std::string &mode) {
         if (fp) {
             close();
         }
@@ -202,60 +221,16 @@ void ClusterFile<ClusterType, Enable>::open(const std::string &mode) {
         } else {
             throw std::runtime_error("Unsupported mode: " + mode);
         }
-}
-
-template <typename ClusterType, typename Enable>
-void ClusterFile<ClusterType, Enable>::set_roi(ROI roi) {
-    m_roi = roi;
-}
-template <typename ClusterType, typename Enable>
-void ClusterFile<ClusterType, Enable>::set_noise_map(
-    const NDView<int32_t, 2> noise_map) {
-    m_noise_map = NDArray<int32_t, 2>(noise_map);
-}
-template <typename ClusterType, typename Enable>
-void ClusterFile<ClusterType, Enable>::set_gain_map(
-    const NDView<double, 2> gain_map) {
-    m_gain_map = GainMap(gain_map);
-}
-
-template <typename ClusterType, typename Enable>
-void ClusterFile<ClusterType, Enable>::set_gain_map(const GainMap &gain_map) {
-    m_gain_map = gain_map;
-}
-
-template <typename ClusterType, typename Enable>
-void ClusterFile<ClusterType, Enable>::set_gain_map(const GainMap &&gain_map) {
-    m_gain_map = gain_map;
-}
-
-// TODO generally supported for all clsuter types
-template <typename ClusterType, typename Enable>
-void ClusterFile<ClusterType, Enable>::write_frame(
-    const ClusterVector<ClusterType> &clusters) {
-    if (m_mode != "w" && m_mode != "a") {
-        throw std::runtime_error("File not opened for writing");
     }
 
-    int32_t frame_number = clusters.frame_number();
+  private:
-    fwrite(&frame_number, sizeof(frame_number), 1, fp);
+    ClusterVector<ClusterType> read_clusters_with_cut(size_t n_clusters);
-    uint32_t n_clusters = clusters.size();
+    ClusterVector<ClusterType> read_clusters_without_cut(size_t n_clusters);
-    fwrite(&n_clusters, sizeof(n_clusters), 1, fp);
+    ClusterVector<ClusterType> read_frame_with_cut();
-    fwrite(clusters.data(), clusters.item_size(), clusters.size(), fp);
+    ClusterVector<ClusterType> read_frame_without_cut();
-}
+    bool is_selected(ClusterType &cl);
-
+    ClusterType read_one_cluster();
-template <typename ClusterType, typename Enable>
+};
-ClusterVector<ClusterType>
-ClusterFile<ClusterType, Enable>::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);
-    }
-}
 
 template <typename ClusterType, typename Enable>
 ClusterVector<ClusterType>
@@ -276,8 +251,8 @@ ClusterFile<ClusterType, Enable>::read_clusters_without_cut(size_t n_clusters) {
     // 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
+            // if we have more photons left in the frame then photons to
-            // read directly the requested number
+            // read we read directly the requested number
             nn = n_clusters;
         } else {
             nn = nph;
@@ -343,8 +318,8 @@ ClusterFile<ClusterType, Enable>::read_clusters_with_cut(size_t n_clusters) {
         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
+                    frame_number); // cluster vector will hold the last
-                                   // number
+                                   // frame number
                 while (m_num_left && clusters.size() < n_clusters) {
                     ClusterType c = read_one_cluster();
                     if (is_selected(c)) {
@@ -375,18 +350,6 @@ ClusterType ClusterFile<ClusterType, Enable>::read_one_cluster() {
     return c;
 }
 
-template <typename ClusterType, typename Enable>
-ClusterVector<ClusterType> ClusterFile<ClusterType, Enable>::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();
-    }
-}
-
 template <typename ClusterType, typename Enable>
 ClusterVector<ClusterType>
 ClusterFile<ClusterType, Enable>::read_frame_without_cut() {
@@ -408,11 +371,15 @@ ClusterFile<ClusterType, Enable>::read_frame_without_cut() {
                                  "Could not read number of clusters");
     }
 
+    LOG(logDEBUG1) << "Reading " << n_clusters << " clusters from frame "
+                   << frame_number;
+
     ClusterVector<ClusterType> clusters(n_clusters);
     clusters.set_frame_number(frame_number);
-
     clusters.resize(n_clusters);
 
+    LOG(logDEBUG1) << "clusters.item_size(): " << clusters.item_size();
+
     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");
@@ -465,13 +432,9 @@ bool ClusterFile<ClusterType, Enable>::is_selected(ClusterType &cl) {
         }
     }
 
-    auto cluster_size_x = extract_template_arguments<
-        std::remove_reference_t<decltype(cl)>>::cluster_size_x;
-    auto cluster_size_y = extract_template_arguments<
-        std::remove_reference_t<decltype(cl)>>::cluster_size_y;
-
     size_t cluster_center_index =
-        (cluster_size_x / 2) + (cluster_size_y / 2) * cluster_size_x;
+        (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

include/aare/ClusterFileSink.hpp

@@ -21,7 +21,7 @@ class ClusterFileSink {
 
     void process() {
         m_stopped = false;
-        fmt::print("ClusterFileSink started\n");
+        LOG(logDEBUG) << "ClusterFileSink started";
         while (!m_stop_requested || !m_source->isEmpty()) {
             if (ClusterVector<ClusterType> *clusters = m_source->frontPtr();
                 clusters != nullptr) {
@@ -41,13 +41,16 @@ class ClusterFileSink {
                 std::this_thread::sleep_for(m_default_wait);
             }
         }
-        fmt::print("ClusterFileSink stopped\n");
+        LOG(logDEBUG) << "ClusterFileSink stopped";
         m_stopped = true;
     }
 
   public:
     ClusterFileSink(ClusterFinderMT<ClusterType, uint16_t, double> *source,
                     const std::filesystem::path &fname) {
+        LOG(logDEBUG) << "ClusterFileSink: "
+                      << "source: " << source->sink()
+                      << ", file: " << fname.string();
         m_source = source->sink();
         m_thread = std::thread(&ClusterFileSink::process, this);
         m_file.open(fname, std::ios::binary);

include/aare/ClusterFinder.hpp

@@ -20,11 +20,9 @@ class ClusterFinder {
     Pedestal<PEDESTAL_TYPE> m_pedestal;
     ClusterVector<ClusterType> m_clusters;
 
-    static const uint8_t ClusterSizeX =
+    static const uint8_t ClusterSizeX = ClusterType::cluster_size_x;
-        extract_template_arguments<ClusterType>::cluster_size_x;
+    static const uint8_t ClusterSizeY = ClusterType::cluster_size_y;
-    static const uint8_t ClusterSizeY =
+    using CT = typename ClusterType::value_type;
-        extract_template_arguments<ClusterType>::cluster_size_x;
-    using CT = typename extract_template_arguments<ClusterType>::value_type;
 
   public:
     /**
@@ -40,7 +38,11 @@ class ClusterFinder {
         : 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) {};
+          m_pedestal(image_size[0], image_size[1]), m_clusters(capacity) {
+        LOG(logDEBUG) << "ClusterFinder: "
+                      << "image_size: " << image_size[0] << "x" << image_size[1]
+                      << ", nSigma: " << nSigma << ", capacity: " << capacity;
+    }
 
     void push_pedestal_frame(NDView<FRAME_TYPE, 2> frame) {
         m_pedestal.push(frame);
@@ -79,7 +81,6 @@ class ClusterFinder {
         int has_center_pixel_y = ClusterSizeY % 2;
 
         m_clusters.set_frame_number(frame_number);
-        std::vector<CT> cluster_data(ClusterSizeX * ClusterSizeY);
         for (int iy = 0; iy < frame.shape(0); iy++) {
             for (int ix = 0; ix < frame.shape(1); ix++) {
 
@@ -126,8 +127,9 @@ class ClusterFinder {
 
                 // Store cluster
                 if (value == max) {
-                    // Zero out the cluster data
+                    ClusterType cluster{};
-                    std::fill(cluster_data.begin(), cluster_data.end(), 0);
+                    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
@@ -141,20 +143,15 @@ class ClusterFinder {
                                     static_cast<CT>(frame(iy + ir, ix + ic)) -
                                     static_cast<CT>(
                                         m_pedestal.mean(iy + ir, ix + ic));
-                                cluster_data[i] =
+                                cluster.data[i] =
                                     tmp; // Watch for out of bounds access
                                 i++;
                             }
                         }
                     }
 
-                    ClusterType new_cluster{};
-                    new_cluster.x = ix;
-                    new_cluster.y = iy;
-                    std::copy(cluster_data.begin(), cluster_data.end(),
-                              new_cluster.data);
                     // Add the cluster to the output ClusterVector
-                    m_clusters.push_back(new_cluster);
+                    m_clusters.push_back(cluster);
                 }
             }
         }

include/aare/ClusterFinderMT.hpp

@@ -8,6 +8,7 @@
 #include "aare/ClusterFinder.hpp"
 #include "aare/NDArray.hpp"
 #include "aare/ProducerConsumerQueue.hpp"
+#include "aare/logger.hpp"
 
 namespace aare {
 
@@ -34,7 +35,8 @@ template <typename ClusterType = Cluster<int32_t, 3, 3>,
           typename FRAME_TYPE = uint16_t, typename PEDESTAL_TYPE = double>
 class ClusterFinderMT {
 
-    using CT = typename extract_template_arguments<ClusterType>::value_type;
+  protected:
+    using CT = typename ClusterType::value_type;
     size_t m_current_thread{0};
     size_t m_n_threads{0};
     using Finder = ClusterFinder<ClusterType, FRAME_TYPE, PEDESTAL_TYPE>;
@@ -50,6 +52,7 @@ 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};
 
@@ -120,6 +123,13 @@ class ClusterFinderMT {
     ClusterFinderMT(Shape<2> image_size, PEDESTAL_TYPE nSigma = 5.0,
                     size_t capacity = 2000, size_t n_threads = 3)
         : m_n_threads(n_threads) {
+
+        LOG(logDEBUG1) << "ClusterFinderMT: "
+                       << "image_size: " << image_size[0] << "x"
+                       << image_size[1] << ", nSigma: " << nSigma
+                       << ", capacity: " << capacity
+                       << ", n_threads: " << n_threads;
+
         for (size_t i = 0; i < n_threads; i++) {
             m_cluster_finders.push_back(
                 std::make_unique<

include/aare/ClusterVector.hpp

@@ -32,7 +32,8 @@ class ClusterVector; // Forward declaration
  */
 template <typename T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
           typename CoordType>
-class ClusterVector<Cluster<T, ClusterSizeX, ClusterSizeY, 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
@@ -47,7 +48,7 @@ class ClusterVector<Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>> {
      * @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 capacity = 300, uint64_t frame_number = 0)
+    ClusterVector(size_t capacity = 1024, uint64_t frame_number = 0)
         : m_frame_number(frame_number) {
         m_data.reserve(capacity);
     }
@@ -76,9 +77,10 @@ class ClusterVector<Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>> {
     std::vector<T> sum() {
         std::vector<T> sums(m_data.size());
 
-        for (size_t i = 0; i < m_data.size(); i++) {
+        std::transform(
-            sums[i] = at(i).sum();
+            m_data.begin(), m_data.end(), sums.begin(),
-        }
+            [](const ClusterType &cluster) { return cluster.sum(); });
+
         return sums;
     }
 
@@ -86,13 +88,15 @@ class ClusterVector<Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>> {
      * @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
-     */ //TODO if underlying container is a vector use std::for_each
+     */
     std::vector<T> sum_2x2() {
         std::vector<T> sums_2x2(m_data.size());
 
-        for (size_t i = 0; i < m_data.size(); i++) {
+        std::transform(m_data.begin(), m_data.end(), sums_2x2.begin(),
-            sums_2x2[i] = at(i).max_sum_2x2().first;
+                       [](const ClusterType &cluster) {
-        }
+                           return cluster.max_sum_2x2().first;
+                       });
+
         return sums_2x2;
     }
 
@@ -119,6 +123,11 @@ class ClusterVector<Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>> {
      */
     size_t size() const { return m_data.size(); }
 
+    /**
+     * @brief Check if the vector is empty
+     */
+    bool empty() const { return m_data.empty(); }
+
     uint8_t cluster_size_x() const { return ClusterSizeX; }
 
     uint8_t cluster_size_y() const { return ClusterSizeY; }
@@ -130,9 +139,9 @@ class ClusterVector<Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>> {
      */
     size_t capacity() const { return m_data.capacity(); }
 
-    const auto begin() const { return m_data.begin(); }
+    auto begin() const { return m_data.begin(); }
 
-    const auto end() const { return m_data.end(); }
+    auto end() const { return m_data.end(); }
 
     /**
      * @brief Return the size in bytes of a single cluster
@@ -149,9 +158,9 @@ class ClusterVector<Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>> {
      * @brief Return a reference to the i-th cluster casted to type V
      * @tparam V type of the cluster
      */
-    ClusterType &at(size_t i) { return m_data[i]; }
+    ClusterType &operator[](size_t i) { return m_data[i]; }
 
-    const ClusterType &at(size_t i) const { return m_data[i]; }
+    const ClusterType &operator[](size_t i) const { return m_data[i]; }
 
     /**
      * @brief Return the frame number of the clusters. 0 is used to indicate
@@ -164,4 +173,22 @@ class ClusterVector<Cluster<T, ClusterSizeX, ClusterSizeY, CoordType>> {
     }
 };
 
+template<typename T>
+ClusterVector<Cluster<T, 2, 2, uint16_t>> reduce_3x3_to_2x2(const ClusterVector<Cluster<T, 3, 3, uint16_t>> &cv) {
+    ClusterVector<Cluster<T, 2, 2, uint16_t>> result;
+    for (const auto &c : cv) {
+        result.push_back(reduce_3x3_to_2x2(c));
+    }
+    return result;
+}
+
+template<typename T>
+ClusterVector<Cluster<T, 3, 3, uint16_t>> reduce_5x5_to_3x3(const ClusterVector<Cluster<T, 5, 5, uint16_t>> &cv) {
+    ClusterVector<Cluster<T, 3, 3, uint16_t>> result;
+    for (const auto &c : cv) {
+        result.push_back(reduce_5x5_to_3x3(c));
+    }
+    return result;
+}
+
 } // namespace aare

include/aare/CtbRawFile.hpp

@@ -1,25 +1,25 @@
 #pragma once
 
 #include "aare/FileInterface.hpp"
-#include "aare/RawMasterFile.hpp"
 #include "aare/Frame.hpp"
+#include "aare/RawMasterFile.hpp"
 
 #include <filesystem>
 #include <fstream>
 
-namespace aare{
+namespace aare {
 
-
+class CtbRawFile {
-class CtbRawFile{
     RawMasterFile m_master;
     std::ifstream m_file;
     size_t m_current_frame{0};
     size_t m_current_subfile{0};
     size_t m_num_subfiles{0};
-public:
+
+  public:
     CtbRawFile(const std::filesystem::path &fname);
 
-    void read_into(std::byte *image_buf, DetectorHeader* header = nullptr);
+    void read_into(std::byte *image_buf, DetectorHeader *header = nullptr);
     void seek(size_t frame_index); //!< seek to the given frame index
     size_t tell() const;           //!< get the frame index of the file pointer
 
@@ -29,13 +29,13 @@ public:
     size_t frames_in_file() const;
 
     RawMasterFile master() const;
-private:
+
+  private:
     void find_subfiles();
     size_t sub_file_index(size_t frame_index) const {
         return frame_index / m_master.max_frames_per_file();
     }
     void open_data_file(size_t subfile_index);
-
 };
 
-}
+} // namespace aare

include/aare/DetectorGeometry.hpp

@@ -0,0 +1,81 @@
+#pragma once
+#include "aare/RawMasterFile.hpp" //ROI refactor away
+#include "aare/defs.hpp"
+namespace aare {
+
+struct ModuleConfig {
+    int module_gap_row{};
+    int module_gap_col{};
+
+    bool operator==(const ModuleConfig &other) const {
+        if (module_gap_col != other.module_gap_col)
+            return false;
+        if (module_gap_row != other.module_gap_row)
+            return false;
+        return true;
+    }
+};
+
+/**
+ * @brief Class to hold the geometry of a module. Where pixel 0 is located and
+ * the size of the module
+ */
+struct ModuleGeometry {
+    int origin_x{};
+    int origin_y{};
+    int height{};
+    int width{};
+    int row_index{};
+    int col_index{};
+};
+
+/**
+ * @brief Class to hold the geometry of a detector. Number of modules, their
+ * size and where pixel 0 for each module is located
+ */
+class DetectorGeometry {
+  public:
+    DetectorGeometry(const xy &geometry, const ssize_t module_pixels_x,
+                     const ssize_t module_pixels_y,
+                     const xy udp_interfaces_per_module = xy{1, 1},
+                     const bool quad = false);
+
+    ~DetectorGeometry() = default;
+
+    /**
+     * @brief Update the detector geometry given a region of interest
+     *
+     * @param roi
+     * @return DetectorGeometry
+     */
+    void update_geometry_with_roi(ROI roi);
+
+    size_t n_modules() const;
+
+    size_t n_modules_in_roi() const;
+
+    size_t pixels_x() const;
+    size_t pixels_y() const;
+
+    size_t modules_x() const;
+    size_t modules_y() const;
+
+    const std::vector<ssize_t> &get_modules_in_roi() const;
+
+    ssize_t get_modules_in_roi(const size_t index) const;
+
+    const std::vector<ModuleGeometry> &get_module_geometries() const;
+
+    const ModuleGeometry &get_module_geometries(const size_t index) const;
+
+  private:
+    size_t m_modules_x{};
+    size_t m_modules_y{};
+    size_t m_pixels_x{};
+    size_t m_pixels_y{};
+    static constexpr ModuleConfig cfg{0, 0};
+    std::vector<ModuleGeometry> module_geometries{};
+    std::vector<ssize_t> modules_in_roi{};
+};
+
+} // namespace aare

include/aare/Dtype.hpp

@@ -6,31 +6,37 @@
 
 namespace aare {
 
-// The format descriptor is a single character that specifies the type of the data
+// The format descriptor is a single character that specifies the type of the
+// data
 // - python documentation: https://docs.python.org/3/c-api/arg.html#numbers
-// - py::format_descriptor<T>::format() (in pybind11) does not return the same format as
+// - py::format_descriptor<T>::format() (in pybind11) does not return the same
+// format as
 //   written in python.org documentation.
-// - numpy also doesn't use the same format. and also numpy associates the format
+// - numpy also doesn't use the same format. and also numpy associates the
-//   with variable bitdepth types. (e.g. long is int64 on linux64 and int32 on win64)
+// format
-//   https://numpy.org/doc/stable/reference/arrays.scalars.html
+//   with variable bitdepth types. (e.g. long is int64 on linux64 and int32 on
+//   win64) https://numpy.org/doc/stable/reference/arrays.scalars.html
 //
 // github issue discussing this:
 // https://github.com/pybind/pybind11/issues/1908#issuecomment-658358767
 //
-// [IN LINUX] the difference is for int64 (long) and uint64 (unsigned long). The format
+// [IN LINUX] the difference is for int64 (long) and uint64 (unsigned long). The
-// descriptor is 'q' and 'Q' respectively and in the documentation it is 'l' and 'k'.
+// format descriptor is 'q' and 'Q' respectively and in the documentation it is
+// 'l' and 'k'.
 
 // in practice numpy doesn't seem to care when reading buffer info: the library
 // interprets 'q' or 'l' as int64 and 'Q' or 'L' as uint64.
-// for this reason we decided to use the same format descriptor as pybind to avoid
+// for this reason we decided to use the same format descriptor as pybind to
-// any further discrepancies.
+// avoid any further discrepancies.
 
 // in the following order:
 // int8, uint8, int16, uint16, int32, uint32, int64, uint64, float, double
-const char DTYPE_FORMAT_DSC[] = {'b', 'B', 'h', 'H', 'i', 'I', 'q', 'Q', 'f', 'd'};
+const char DTYPE_FORMAT_DSC[] = {'b', 'B', 'h', 'H', 'i',
+                                 'I', 'q', 'Q', 'f', 'd'};
 
 // on linux64 & apple
-const char NUMPY_FORMAT_DSC[] = {'b', 'B', 'h', 'H', 'i', 'I', 'l', 'L', 'f', 'd'};
+const char NUMPY_FORMAT_DSC[] = {'b', 'B', 'h', 'H', 'i',
+                                 'I', 'l', 'L', 'f', 'd'};
 /**
  * @brief enum class to define the endianess of the system
  */
@@ -52,12 +58,29 @@ enum class endian {
  */
 class Dtype {
   public:
-    enum TypeIndex { INT8, UINT8, INT16, UINT16, INT32, UINT32, INT64, UINT64, FLOAT, DOUBLE, ERROR, NONE };
+    enum TypeIndex {
+        INT8,
+        UINT8,
+        INT16,
+        UINT16,
+        INT32,
+        UINT32,
+        INT64,
+        UINT64,
+        FLOAT,
+        DOUBLE,
+        ERROR,
+        NONE
+    };
 
     uint8_t bitdepth() const;
     size_t bytes() const;
-    std::string format_descr() const { return std::string(1, DTYPE_FORMAT_DSC[static_cast<int>(m_type)]); }
+    std::string format_descr() const {
-    std::string numpy_descr() const { return std::string(1, NUMPY_FORMAT_DSC[static_cast<int>(m_type)]); }
+        return std::string(1, DTYPE_FORMAT_DSC[static_cast<int>(m_type)]);
+    }
+    std::string numpy_descr() const {
+        return std::string(1, NUMPY_FORMAT_DSC[static_cast<int>(m_type)]);
+    }
 
     explicit Dtype(const std::type_info &t);
     explicit Dtype(std::string_view sv);

include/aare/File.hpp

@@ -6,11 +6,11 @@ namespace aare {
 
 /**
  * @brief RAII File class for reading, and in the future potentially writing
- * image files in various formats. Minimal generic interface. For specail fuctions
+ * image files in various formats. Minimal generic interface. For specail
- * plase use the RawFile or NumpyFile classes directly.
+ * fuctions plase use the RawFile or NumpyFile classes directly. Wraps
- * Wraps FileInterface to abstract the underlying file format
+ * FileInterface to abstract the underlying file format
- * @note **frame_number** refers the the frame number sent by the detector while **frame_index**
+ * @note **frame_number** refers the the frame number sent by the detector while
- * is the position of the frame in the file
+ * **frame_index** is the position of the frame in the file
  */
 class File {
     std::unique_ptr<FileInterface> file_impl;
@@ -25,29 +25,35 @@ class File {
      * @throws std::invalid_argument if the file mode is not supported
      *
      */
-    File(const std::filesystem::path &fname, const std::string &mode="r", const FileConfig &cfg = {});
+    File(const std::filesystem::path &fname, const std::string &mode = "r",
+         const FileConfig &cfg = {});
 
-     /**Since the object is responsible for managing the file we disable copy construction */
+    /**Since the object is responsible for managing the file we disable copy
+     * construction */
     File(File const &other) = delete;
 
     /**The same goes for copy assignment */
-    File& operator=(File const &other) = delete;
+    File &operator=(File const &other) = delete;
 
     File(File &&other) noexcept;
-    File& operator=(File &&other) noexcept;
+    File &operator=(File &&other) noexcept;
     ~File() = default;
 
     // void close();                             //!< close the file
 
-    Frame read_frame();                       //!< read one frame from the file at the current position
+    Frame
-    Frame read_frame(size_t frame_index);     //!< read one frame at the position given by frame number
+    read_frame(); //!< read one frame from the file at the current position
-    std::vector<Frame> read_n(size_t n_frames); //!< read n_frames from the file at the current position
+    Frame read_frame(size_t frame_index); //!< read one frame at the position
+                                          //!< given by frame number
+    std::vector<Frame> read_n(size_t n_frames); //!< read n_frames from the file
+                                                //!< at the current position
 
     void read_into(std::byte *image_buf);
     void read_into(std::byte *image_buf, size_t n_frames);
 
     size_t frame_number(); //!< get the frame number at the current position
-    size_t frame_number(size_t frame_index);  //!< get the frame number at the given frame index
+    size_t frame_number(
+        size_t frame_index); //!< get the frame number at the given frame index
     size_t bytes_per_frame() const;
     size_t pixels_per_frame() const;
     size_t bytes_per_pixel() const;
@@ -59,8 +65,6 @@ class File {
     size_t cols() const;
 
     DetectorType detector_type() const;
-
-    
 };
 
 } // namespace aare

include/aare/FileInterface.hpp

@@ -20,8 +20,10 @@ struct FileConfig {
     uint64_t rows{};
     uint64_t cols{};
     bool operator==(const FileConfig &other) const {
-        return dtype == other.dtype && rows == other.rows && cols == other.cols && geometry == other.geometry &&
+        return dtype == other.dtype && rows == other.rows &&
-               detector_type == other.detector_type && max_frames_per_file == other.max_frames_per_file;
+               cols == other.cols && geometry == other.geometry &&
+               detector_type == other.detector_type &&
+               max_frames_per_file == other.max_frames_per_file;
     }
     bool operator!=(const FileConfig &other) const { return !(*this == other); }
 
@@ -32,8 +34,11 @@ struct FileConfig {
     int max_frames_per_file{};
     size_t total_frames{};
     std::string to_string() const {
-        return "{ dtype: " + dtype.to_string() + ", rows: " + std::to_string(rows) + ", cols: " + std::to_string(cols) +
+        return "{ dtype: " + dtype.to_string() +
-               ", geometry: " + geometry.to_string() + ", detector_type: " + ToString(detector_type) +
+               ", rows: " + std::to_string(rows) +
+               ", cols: " + std::to_string(cols) +
+               ", geometry: " + geometry.to_string() +
+               ", detector_type: " + ToString(detector_type) +
                ", max_frames_per_file: " + std::to_string(max_frames_per_file) +
                ", total_frames: " + std::to_string(total_frames) + " }";
     }
@@ -42,7 +47,8 @@ struct FileConfig {
 /**
  * @brief FileInterface class to define the interface for file operations
  * @note parent class for NumpyFile and RawFile
- * @note all functions are pure virtual and must be implemented by the derived classes
+ * @note all functions are pure virtual and must be implemented by the derived
+ * classes
  */
 class FileInterface {
   public:
@@ -64,17 +70,20 @@ class FileInterface {
      * @param n_frames number of frames to read
      * @return vector of frames
      */
-    virtual std::vector<Frame> read_n(size_t n_frames) = 0; // Is this the right interface?
+    virtual std::vector<Frame>
+    read_n(size_t n_frames) = 0; // Is this the right interface?
 
     /**
-     * @brief read one frame from the file at the current position and store it in the provided buffer
+     * @brief read one frame from the file at the current position and store it
+     * in the provided buffer
      * @param image_buf buffer to store the frame
      * @return void
      */
     virtual void read_into(std::byte *image_buf) = 0;
 
     /**
-     * @brief read n_frames from the file at the current position and store them in the provided buffer
+     * @brief read n_frames from the file at the current position and store them
+     * in the provided buffer
      * @param image_buf buffer to store the frames
      * @param n_frames number of frames to read
      * @return void
@@ -134,7 +143,6 @@ class FileInterface {
      */
     virtual size_t bitdepth() const = 0;
 
-
     virtual DetectorType detector_type() const = 0;
 
     // function to query the data type of the file

include/aare/FilePtr.hpp

@@ -12,14 +12,14 @@ class FilePtr {
 
   public:
     FilePtr() = default;
-    FilePtr(const std::filesystem::path& fname, const std::string& mode);
+    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();
+    ssize_t tell();
-    void seek(int64_t offset, int whence = SEEK_SET) {
+    void seek(ssize_t offset, int whence = SEEK_SET) {
         if (fseek(fp_, offset, whence) != 0)
             throw std::runtime_error("Error seeking in file");
     }

include/aare/Fit.hpp

@@ -15,15 +15,27 @@ NDArray<double, 1> gaus(NDView<double, 1> x, NDView<double, 1> par);
 double pol1(const double x, const double *par);
 NDArray<double, 1> pol1(NDView<double, 1> x, NDView<double, 1> par);
 
-} // namespace func
+double scurve(const double x, const double *par);
+NDArray<double, 1> scurve(NDView<double, 1> x, NDView<double, 1> par);
 
+double scurve2(const double x, const double *par);
+NDArray<double, 1> scurve2(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, 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);
+std::array<double, 2> pol1_init_par(const NDView<double, 1> x,
+                                    const NDView<double, 1> y);
+
+std::array<double, 6> scurve_init_par(const NDView<double, 1> x,
+                                      const NDView<double, 1> y);
+std::array<double, 6> scurve2_init_par(const NDView<double, 1> x,
+                                       const NDView<double, 1> y);
 
 static constexpr int DEFAULT_NUM_THREADS = 4;
 
@@ -34,46 +46,41 @@ static constexpr int DEFAULT_NUM_THREADS = 4;
  */
 NDArray<double, 1> fit_gaus(NDView<double, 1> x, NDView<double, 1> y);
 
-
 /**
  * @brief Fit a 1D Gaussian to each pixel. Data layout [row, col, values]
  * @param x x values
- * @param y y vales, layout [row, col, values]
+ * @param y y values, 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);
 
-
-
-
 /**
  * @brief Fit a 1D Gaussian with error estimates
  * @param x x values
- * @param y y vales, layout [row, col, values]
+ * @param y y values, layout [row, col, values]
  * @param y_err error in y, layout [row, col, values]
  * @param par_out output parameters
  * @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,
-              double& chi2);
+              double &chi2);
 
 /**
  * @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 y values, layout [row, col, values]
  * @param y_err error in y, layout [row, col, values]
  * @param par_out output parameters, layout [row, col, values]
  * @param par_err_out output parameter errors, layout [row, col, values]
  * @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, NDView<double, 2> chi2_out,
+              NDView<double, 3> par_out, NDView<double, 3> par_err_out,
-              int n_threads = DEFAULT_NUM_THREADS
+              NDView<double, 2> chi2_out, int n_threads = DEFAULT_NUM_THREADS);
-              );
 
 NDArray<double, 1> fit_pol1(NDView<double, 1> x, NDView<double, 1> y);
 
@@ -81,12 +88,33 @@ 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, double& chi2);
+              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,
+              NDView<double, 3> par_out, NDView<double, 3> par_err_out,
-              int n_threads = DEFAULT_NUM_THREADS);
+              NDView<double, 2> chi2_out, int n_threads = DEFAULT_NUM_THREADS);
 
+NDArray<double, 1> fit_scurve(NDView<double, 1> x, NDView<double, 1> y);
+NDArray<double, 3> fit_scurve(NDView<double, 1> x, NDView<double, 3> y,
+                              int n_threads);
+void fit_scurve(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);
+void fit_scurve(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);
 
+NDArray<double, 1> fit_scurve2(NDView<double, 1> x, NDView<double, 1> y);
+NDArray<double, 3> fit_scurve2(NDView<double, 1> x, NDView<double, 3> y,
+                               int n_threads);
+void fit_scurve2(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);
+void fit_scurve2(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);
 } // namespace aare

include/aare/Frame.hpp

@@ -19,7 +19,7 @@ class Frame {
     uint32_t m_cols;
     Dtype m_dtype;
     std::byte *m_data;
-    //TODO! Add frame number?
+    // TODO! Add frame number?
 
   public:
     /**
@@ -39,7 +39,7 @@ class Frame {
      * @param dtype data type of the pixels
      */
     Frame(const std::byte *bytes, uint32_t rows, uint32_t cols, Dtype dtype);
-    ~Frame(){ delete[] m_data; };
+    ~Frame() { delete[] m_data; };
 
     /** @warning Copy is disabled to ensure performance when passing
      * frames around. Can discuss enabling it.
@@ -52,7 +52,6 @@ class Frame {
     Frame &operator=(Frame &&other) noexcept;
     Frame(Frame &&other) noexcept;
 
-
     Frame clone() const; //<- Explicit copy
 
     uint32_t rows() const;
@@ -93,7 +92,7 @@ class Frame {
         if (row >= m_rows || col >= m_cols) {
             throw std::out_of_range("Invalid row or column index");
         }
-        //TODO! add tests then reimplement using pixel_ptr
+        // TODO! add tests then reimplement using pixel_ptr
         T data;
         std::memcpy(&data, m_data + (row * m_cols + col) * m_dtype.bytes(),
                     m_dtype.bytes());
@@ -107,8 +106,8 @@ class Frame {
      * @return NDView<T, 2>
      */
     template <typename T> NDView<T, 2> view() {
-        std::array<int64_t, 2> shape = {static_cast<int64_t>(m_rows),
+        std::array<ssize_t, 2> shape = {static_cast<ssize_t>(m_rows),
-                                        static_cast<int64_t>(m_cols)};
+                                        static_cast<ssize_t>(m_cols)};
         T *data = reinterpret_cast<T *>(m_data);
         return NDView<T, 2>(data, shape);
     }

include/aare/GainMap.hpp

@@ -1,6 +1,7 @@
 /************************************************
- * @file ApplyGainMap.hpp
+ * @file GainMap.hpp
- * @short function to apply gain map of image size to a vector of clusters
+ * @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
@@ -12,14 +13,21 @@
 
 namespace aare {
 
-class GainMap {
+class InvertedGainMap {
 
   public:
-    explicit GainMap(const NDArray<double, 2> &gain_map)
+    explicit InvertedGainMap(const NDArray<double, 2> &gain_map)
-        : m_gain_map(gain_map) {};
+        : m_gain_map(gain_map) {
+        for (auto &item : m_gain_map) {
+            item = 1.0 / item;
+        }
+    };
 
-    explicit GainMap(const NDView<double, 2> gain_map) {
+    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,
@@ -34,19 +42,21 @@ class GainMap {
         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.at(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] = cl.data[j] * static_cast<T>(m_gain_map(y, x));
+                    cl.data[j] = static_cast<T>(
+                        static_cast<double>(cl.data[j]) *
+                        m_gain_map(
+                            y, x)); // cast after conversion to keep precision
                 }
             } else {
-                memset(cl.data, 0,
+                // clear edge clusters
-                       ClusterSizeX * ClusterSizeY *
+                cl.data.fill(0);
-                           sizeof(T)); // clear edge clusters
             }
         }
     }

include/aare/Interpolator.hpp

@@ -44,15 +44,14 @@ Interpolator::interpolate(const ClusterVector<ClusterType> &clusters) {
     photons.reserve(clusters.size());
 
     if (clusters.cluster_size_x() == 3 || clusters.cluster_size_y() == 3) {
-        for (size_t i = 0; i < clusters.size(); i++) {
+        for (const ClusterType &cluster : clusters) {
 
-            auto cluster = clusters.at(i);
             auto eta = calculate_eta2(cluster);
 
             Photon photon;
             photon.x = cluster.x;
             photon.y = cluster.y;
-            photon.energy = eta.sum;
+            photon.energy = static_cast<decltype(photon.energy)>(eta.sum);
 
             // auto ie = nearest_index(m_energy_bins, photon.energy)-1;
             // auto ix = nearest_index(m_etabinsx, eta.x)-1;
@@ -70,20 +69,20 @@ Interpolator::interpolate(const ClusterVector<ClusterType> &clusters) {
             // cBottomRight = 1,
             // cTopLeft = 2,
             // cTopRight = 3
-            switch (eta.c) {
+            switch (static_cast<corner>(eta.c)) {
-            case cTopLeft:
+            case corner::cTopLeft:
                 dX = -1.;
                 dY = 0;
                 break;
-            case cTopRight:;
+            case corner::cTopRight:;
                 dX = 0;
                 dY = 0;
                 break;
-            case cBottomLeft:
+            case corner::cBottomLeft:
                 dX = -1.;
                 dY = -1.;
                 break;
-            case cBottomRight:
+            case corner::cBottomRight:
                 dX = 0.;
                 dY = -1.;
                 break;
@@ -94,14 +93,13 @@ Interpolator::interpolate(const ClusterVector<ClusterType> &clusters) {
         }
     } else if (clusters.cluster_size_x() == 2 ||
                clusters.cluster_size_y() == 2) {
-        for (size_t i = 0; i < clusters.size(); i++) {
+        for (const ClusterType &cluster : clusters) {
-            auto cluster = clusters.at(i);
             auto eta = calculate_eta2(cluster);
 
             Photon photon;
             photon.x = cluster.x;
             photon.y = cluster.y;
-            photon.energy = eta.sum;
+            photon.energy = static_cast<decltype(photon.energy)>(eta.sum);
 
             // Now do some actual interpolation.
             // Find which energy bin the cluster is in

include/aare/JungfrauDataFile.hpp

@@ -3,14 +3,13 @@
 #include <filesystem>
 #include <vector>
 
-#include "aare/FilePtr.hpp"
-#include "aare/defs.hpp"
-#include "aare/NDArray.hpp"
 #include "aare/FileInterface.hpp"
+#include "aare/FilePtr.hpp"
+#include "aare/NDArray.hpp"
+#include "aare/defs.hpp"
 namespace aare {
 
-
+struct JungfrauDataHeader {
-struct JungfrauDataHeader{
     uint64_t framenum;
     uint64_t bunchid;
 };
@@ -18,44 +17,50 @@ struct JungfrauDataHeader{
 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_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_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)
+    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::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
+    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)
+    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)
+    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;
+    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;
+    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;
@@ -63,44 +68,48 @@ class JungfrauDataFile : public FileInterface {
 
     /**
      * @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 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 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);
+    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);
+    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); }
+    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)
+     * @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
+     * @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

@@ -21,11 +21,10 @@ TODO! Add expression templates for operators
 
 namespace aare {
 
-
+template <typename T, ssize_t Ndim = 2>
-template <typename T, int64_t Ndim = 2>
 class NDArray : public ArrayExpr<NDArray<T, Ndim>, Ndim> {
-    std::array<int64_t, Ndim> shape_;
+    std::array<ssize_t, Ndim> shape_;
-    std::array<int64_t, Ndim> strides_;
+    std::array<ssize_t, Ndim> strides_;
     size_t size_{};
     T *data_;
 
@@ -34,7 +33,7 @@ class NDArray : public ArrayExpr<NDArray<T, Ndim>, Ndim> {
      * @brief Default constructor. Will construct an empty NDArray.
      *
      */
-    NDArray() : shape_(), strides_(c_strides<Ndim>(shape_)), data_(nullptr) {};
+    NDArray() : shape_(), strides_(c_strides<Ndim>(shape_)), data_(nullptr){};
 
     /**
      * @brief Construct a new NDArray object with a given shape.
@@ -42,20 +41,19 @@ class NDArray : public ArrayExpr<NDArray<T, Ndim>, Ndim> {
      *
      * @param shape shape of the new NDArray
      */
-    explicit NDArray(std::array<int64_t, Ndim> shape)
+    explicit NDArray(std::array<ssize_t, Ndim> shape)
         : shape_(shape), strides_(c_strides<Ndim>(shape_)),
           size_(std::accumulate(shape_.begin(), shape_.end(), 1,
                                 std::multiplies<>())),
           data_(new T[size_]) {}
 
-
     /**
      * @brief Construct a new NDArray object with a shape and value.
      *
      * @param shape shape of the new array
      * @param value value to initialize the array with
      */
-    NDArray(std::array<int64_t, Ndim> shape, T value) : NDArray(shape) {
+    NDArray(std::array<ssize_t, Ndim> shape, T value) : NDArray(shape) {
         this->operator=(value);
     }
 
@@ -69,8 +67,8 @@ class NDArray : public ArrayExpr<NDArray<T, Ndim>, Ndim> {
         std::copy(v.begin(), v.end(), begin());
     }
 
-    template<size_t Size>
+    template <size_t Size>
-    NDArray(const std::array<T, Size>& arr) : NDArray<T,1>({Size}) {
+    NDArray(const std::array<T, Size> &arr) : NDArray<T, 1>({Size}) {
         std::copy(arr.begin(), arr.end(), begin());
     }
 
@@ -79,7 +77,6 @@ class NDArray : public ArrayExpr<NDArray<T, Ndim>, Ndim> {
         : shape_(other.shape_), strides_(c_strides<Ndim>(shape_)),
           size_(other.size_), data_(other.data_) {
         other.reset(); // TODO! is this necessary?
-
     }
 
     // Copy constructor
@@ -113,10 +110,10 @@ 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
+    // Write directly to the data array, or create a new one
-    template<size_t Size>
+    template <size_t Size>
-    NDArray<T,1>& operator=(const std::array<T,Size> &other){
+    NDArray<T, 1> &operator=(const std::array<T, Size> &other) {
-        if(Size != size_){
+        if (Size != size_) {
             delete[] data_;
             size_ = Size;
             data_ = new T[size_];
@@ -157,11 +154,6 @@ 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]);
@@ -186,22 +178,22 @@ class NDArray : public ArrayExpr<NDArray<T, Ndim>, Ndim> {
     }
 
     // TODO! is int the right type for index?
-    T &operator()(int64_t i) { return data_[i]; }
+    T &operator()(ssize_t i) { return data_[i]; }
-    const T &operator()(int64_t i) const { return data_[i]; }
+    const T &operator()(ssize_t i) const { return data_[i]; }
 
-    T &operator[](int64_t i) { return data_[i]; }
+    T &operator[](ssize_t i) { return data_[i]; }
-    const T &operator[](int64_t i) const { return data_[i]; }
+    const T &operator[](ssize_t i) const { return data_[i]; }
 
     T *data() { return data_; }
     std::byte *buffer() { return reinterpret_cast<std::byte *>(data_); }
     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_; }
+    std::array<ssize_t, Ndim> shape() const noexcept { return shape_; }
-    int64_t shape(int64_t i) const noexcept { return shape_[i]; }
+    ssize_t shape(ssize_t i) const noexcept { return shape_[i]; }
-    std::array<int64_t, Ndim> strides() const noexcept { return strides_; }
+    std::array<ssize_t, Ndim> strides() const noexcept { return strides_; }
     size_t bitdepth() const noexcept { return sizeof(T) * 8; }
 
-    std::array<int64_t, Ndim> byte_strides() const noexcept {
+    std::array<ssize_t, Ndim> byte_strides() const noexcept {
         auto byte_strides = strides_;
         for (auto &val : byte_strides)
             val *= sizeof(T);
@@ -228,7 +220,7 @@ class NDArray : public ArrayExpr<NDArray<T, Ndim>, Ndim> {
 };
 
 // Move assign
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> &
 NDArray<T, Ndim>::operator=(NDArray<T, Ndim> &&other) noexcept {
     if (this != &other) {
@@ -242,7 +234,7 @@ NDArray<T, Ndim>::operator=(NDArray<T, Ndim> &&other) noexcept {
     return *this;
 }
 
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> &NDArray<T, Ndim>::operator+=(const NDArray<T, Ndim> &other) {
     // check shape
     if (shape_ == other.shape_) {
@@ -254,7 +246,7 @@ NDArray<T, Ndim> &NDArray<T, Ndim>::operator+=(const NDArray<T, Ndim> &other) {
     throw(std::runtime_error("Shape of ImageDatas must match"));
 }
 
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> &NDArray<T, Ndim>::operator-=(const NDArray<T, Ndim> &other) {
     // check shape
     if (shape_ == other.shape_) {
@@ -266,7 +258,7 @@ NDArray<T, Ndim> &NDArray<T, Ndim>::operator-=(const NDArray<T, Ndim> &other) {
     throw(std::runtime_error("Shape of ImageDatas must match"));
 }
 
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> &NDArray<T, Ndim>::operator*=(const NDArray<T, Ndim> &other) {
     // check shape
     if (shape_ == other.shape_) {
@@ -278,14 +270,14 @@ NDArray<T, Ndim> &NDArray<T, Ndim>::operator*=(const NDArray<T, Ndim> &other) {
     throw(std::runtime_error("Shape of ImageDatas must match"));
 }
 
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> &NDArray<T, Ndim>::operator&=(const T &mask) {
     for (auto it = begin(); it != end(); ++it)
         *it &= mask;
     return *this;
 }
 
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<bool, Ndim> NDArray<T, Ndim>::operator>(const NDArray &other) {
     if (shape_ == other.shape_) {
         NDArray<bool, Ndim> result{shape_};
@@ -297,7 +289,7 @@ NDArray<bool, Ndim> NDArray<T, Ndim>::operator>(const NDArray &other) {
     throw(std::runtime_error("Shape of ImageDatas must match"));
 }
 
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> &NDArray<T, Ndim>::operator=(const NDArray<T, Ndim> &other) {
     if (this != &other) {
         delete[] data_;
@@ -310,7 +302,7 @@ NDArray<T, Ndim> &NDArray<T, Ndim>::operator=(const NDArray<T, Ndim> &other) {
     return *this;
 }
 
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 bool NDArray<T, Ndim>::operator==(const NDArray<T, Ndim> &other) const {
     if (shape_ != other.shape_)
         return false;
@@ -322,83 +314,80 @@ bool NDArray<T, Ndim>::operator==(const NDArray<T, Ndim> &other) const {
     return true;
 }
 
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 bool NDArray<T, Ndim>::operator!=(const NDArray<T, Ndim> &other) const {
     return !((*this) == other);
 }
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> &NDArray<T, Ndim>::operator++() {
     for (uint32_t i = 0; i < size_; ++i)
         data_[i] += 1;
     return *this;
 }
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> &NDArray<T, Ndim>::operator=(const T &value) {
     std::fill_n(data_, size_, value);
     return *this;
 }
 
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> &NDArray<T, Ndim>::operator+=(const T &value) {
     for (uint32_t i = 0; i < size_; ++i)
         data_[i] += value;
     return *this;
 }
 
-
+template <typename T, ssize_t Ndim>
-
-
-template <typename T, int64_t Ndim>
 NDArray<T, Ndim> NDArray<T, Ndim>::operator+(const T &value) {
     NDArray result = *this;
     result += value;
     return result;
 }
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> &NDArray<T, Ndim>::operator-=(const T &value) {
     for (uint32_t i = 0; i < size_; ++i)
         data_[i] -= value;
     return *this;
 }
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> NDArray<T, Ndim>::operator-(const T &value) {
     NDArray result = *this;
     result -= value;
     return result;
 }
 
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> &NDArray<T, Ndim>::operator/=(const T &value) {
     for (uint32_t i = 0; i < size_; ++i)
         data_[i] /= value;
     return *this;
 }
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> NDArray<T, Ndim>::operator/(const T &value) {
     NDArray result = *this;
     result /= value;
     return result;
 }
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> &NDArray<T, Ndim>::operator*=(const T &value) {
     for (uint32_t i = 0; i < size_; ++i)
         data_[i] *= value;
     return *this;
 }
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> NDArray<T, Ndim>::operator*(const T &value) {
     NDArray result = *this;
     result *= value;
     return result;
 }
-// template <typename T, int64_t Ndim> void NDArray<T, Ndim>::Print() {
+// template <typename T, ssize_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>
+template <typename T, ssize_t Ndim>
 std::ostream &operator<<(std::ostream &os, const NDArray<T, Ndim> &arr) {
     for (auto row = 0; row < arr.shape(0); ++row) {
         for (auto col = 0; col < arr.shape(1); ++col) {
@@ -410,7 +399,7 @@ std::ostream &operator<<(std::ostream &os, const NDArray<T, Ndim> &arr) {
     return os;
 }
 
-template <typename T, int64_t Ndim> void NDArray<T, Ndim>::Print_all() {
+template <typename T, ssize_t Ndim> void NDArray<T, Ndim>::Print_all() {
     for (auto row = 0; row < shape_[0]; ++row) {
         for (auto col = 0; col < shape_[1]; ++col) {
             std::cout << std::setw(3);
@@ -419,7 +408,7 @@ template <typename T, int64_t Ndim> void NDArray<T, Ndim>::Print_all() {
         std::cout << "\n";
     }
 }
-template <typename T, int64_t Ndim> void NDArray<T, Ndim>::Print_some() {
+template <typename T, ssize_t Ndim> void NDArray<T, Ndim>::Print_some() {
     for (auto row = 0; row < 5; ++row) {
         for (auto col = 0; col < 5; ++col) {
             std::cout << std::setw(7);
@@ -429,7 +418,7 @@ template <typename T, int64_t Ndim> void NDArray<T, Ndim>::Print_some() {
     }
 }
 
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 void save(NDArray<T, Ndim> &img, std::string &pathname) {
     std::ofstream f;
     f.open(pathname, std::ios::binary);
@@ -437,9 +426,9 @@ void save(NDArray<T, Ndim> &img, std::string &pathname) {
     f.close();
 }
 
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 NDArray<T, Ndim> load(const std::string &pathname,
-                      std::array<int64_t, Ndim> shape) {
+                      std::array<ssize_t, Ndim> shape) {
     NDArray<T, Ndim> img{shape};
     std::ifstream f;
     f.open(pathname, std::ios::binary);
@@ -448,6 +437,4 @@ NDArray<T, Ndim> load(const std::string &pathname,
     return img;
 }
 
-
-
 } // namespace aare

include/aare/NDView.hpp

@@ -1,6 +1,6 @@
 #pragma once
-#include "aare/defs.hpp"
 #include "aare/ArrayExpr.hpp"
+#include "aare/defs.hpp"
 
 #include <algorithm>
 #include <array>
@@ -14,10 +14,11 @@
 #include <vector>
 namespace aare {
 
-template <int64_t Ndim> using Shape = std::array<int64_t, Ndim>;
+template <ssize_t Ndim> using Shape = std::array<ssize_t, Ndim>;
 
 // TODO! fix mismatch between signed and unsigned
-template <int64_t Ndim> Shape<Ndim> make_shape(const std::vector<size_t> &shape) {
+template <ssize_t Ndim>
+Shape<Ndim> make_shape(const std::vector<size_t> &shape) {
     if (shape.size() != Ndim)
         throw std::runtime_error("Shape size mismatch");
     Shape<Ndim> arr;
@@ -25,67 +26,76 @@ template <int64_t Ndim> Shape<Ndim> make_shape(const std::vector<size_t> &shape)
     return arr;
 }
 
-template <int64_t Dim = 0, typename Strides> int64_t element_offset(const Strides & /*unused*/) { return 0; }
+template <ssize_t Dim = 0, typename Strides>
+ssize_t element_offset(const Strides & /*unused*/) {
+    return 0;
+}
 
-template <int64_t Dim = 0, typename Strides, typename... Ix>
+template <ssize_t Dim = 0, typename Strides, typename... Ix>
-int64_t element_offset(const Strides &strides, int64_t i, Ix... index) {
+ssize_t element_offset(const Strides &strides, ssize_t i, Ix... index) {
     return i * strides[Dim] + element_offset<Dim + 1>(strides, index...);
 }
 
-template <int64_t Ndim> std::array<int64_t, Ndim> c_strides(const std::array<int64_t, Ndim> &shape) {
+template <ssize_t Ndim>
-    std::array<int64_t, Ndim> strides{};
+std::array<ssize_t, Ndim> c_strides(const std::array<ssize_t, Ndim> &shape) {
+    std::array<ssize_t, Ndim> strides{};
     std::fill(strides.begin(), strides.end(), 1);
-    for (int64_t i = Ndim - 1; i > 0; --i) {
+    for (ssize_t i = Ndim - 1; i > 0; --i) {
         strides[i - 1] = strides[i] * shape[i];
     }
     return strides;
 }
 
-template <int64_t Ndim> std::array<int64_t, Ndim> make_array(const std::vector<int64_t> &vec) {
+template <ssize_t Ndim>
+std::array<ssize_t, Ndim> make_array(const std::vector<ssize_t> &vec) {
     assert(vec.size() == Ndim);
-    std::array<int64_t, Ndim> arr{};
+    std::array<ssize_t, Ndim> arr{};
     std::copy_n(vec.begin(), Ndim, arr.begin());
     return arr;
 }
 
-template <typename T, int64_t Ndim = 2> class NDView : public ArrayExpr<NDView<T, Ndim>, Ndim> {
+template <typename T, ssize_t Ndim = 2>
+class NDView : public ArrayExpr<NDView<T, Ndim>, Ndim> {
   public:
     NDView() = default;
     ~NDView() = default;
     NDView(const NDView &) = default;
     NDView(NDView &&) = default;
 
-    NDView(T *buffer, std::array<int64_t, Ndim> shape)
+    NDView(T *buffer, std::array<ssize_t, Ndim> shape)
         : buffer_(buffer), strides_(c_strides<Ndim>(shape)), shape_(shape),
-          size_(std::accumulate(std::begin(shape), std::end(shape), 1, std::multiplies<>())) {}
+          size_(std::accumulate(std::begin(shape), std::end(shape), 1,
+                                std::multiplies<>())) {}
 
-    // NDView(T *buffer, const std::vector<int64_t> &shape)
+    template <typename... Ix>
-    //     : buffer_(buffer), strides_(c_strides<Ndim>(make_array<Ndim>(shape))), shape_(make_array<Ndim>(shape)),
+    std::enable_if_t<sizeof...(Ix) == Ndim, T &> operator()(Ix... index) {
-    //       size_(std::accumulate(std::begin(shape), std::end(shape), 1, std::multiplies<>())) {}
-
-    template <typename... Ix> std::enable_if_t<sizeof...(Ix) == Ndim, T &> operator()(Ix... index) {
         return buffer_[element_offset(strides_, index...)];
     }
 
-    template <typename... Ix> std::enable_if_t<sizeof...(Ix) == Ndim, T &> operator()(Ix... index) const {
+    template <typename... Ix>
+    const std::enable_if_t<sizeof...(Ix) == Ndim, T &> operator()(Ix... index) const {
         return buffer_[element_offset(strides_, index...)];
     }
 
     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_; }
+    std::array<ssize_t, Ndim> strides() const noexcept { return strides_; }
 
     T *begin() { return buffer_; }
     T *end() { return buffer_ + size_; }
     T const *begin() const { return buffer_; }
     T const *end() const { return buffer_ + size_; }
-    T &operator()(int64_t i) const { return buffer_[i]; }
+    T &operator()(ssize_t i)  { return buffer_[i]; }
-    T &operator[](int64_t i) const { return buffer_[i]; }
+    T &operator[](ssize_t i)  { return buffer_[i]; }
+    const T &operator()(ssize_t i) const { return buffer_[i]; }
+    const T &operator[](ssize_t i) const { return buffer_[i]; }
 
     bool operator==(const NDView &other) const {
         if (size_ != other.size_)
             return false;
-        for (uint64_t i = 0; i != size_; ++i) {
+        if (shape_ != other.shape_)
+            return false;
+        for (size_t i = 0; i != size_; ++i) {
             if (buffer_[i] != other.buffer_[i])
                 return false;
         }
@@ -94,16 +104,21 @@ template <typename T, int64_t Ndim = 2> class NDView : public ArrayExpr<NDView<T
 
     NDView &operator+=(const T val) { return elemenwise(val, std::plus<T>()); }
     NDView &operator-=(const T val) { return elemenwise(val, std::minus<T>()); }
-    NDView &operator*=(const T val) { return elemenwise(val, std::multiplies<T>()); }
+    NDView &operator*=(const T val) {
-    NDView &operator/=(const T val) { return elemenwise(val, std::divides<T>()); }
+        return elemenwise(val, std::multiplies<T>());
+    }
+    NDView &operator/=(const T val) {
+        return elemenwise(val, std::divides<T>());
+    }
 
-    NDView &operator/=(const NDView &other) { return elemenwise(other, std::divides<T>()); }
+    NDView &operator/=(const NDView &other) {
+        return elemenwise(other, std::divides<T>());
+    }
 
-
+    template <size_t Size> NDView &operator=(const std::array<T, Size> &arr) {
-    template<size_t Size>
+        if (size() != static_cast<ssize_t>(arr.size()))
-    NDView& operator=(const std::array<T, Size> &arr) {
+            throw std::runtime_error(LOCATION +
-        if(size() != static_cast<ssize_t>(arr.size()))
+                                     "Array and NDView size mismatch");
-            throw std::runtime_error(LOCATION + "Array and NDView size mismatch");
         std::copy(arr.begin(), arr.end(), begin());
         return *this;
     }
@@ -136,31 +151,35 @@ template <typename T, int64_t Ndim = 2> class NDView : public ArrayExpr<NDView<T
     }
 
     auto &shape() const { return shape_; }
-    auto shape(int64_t i) const { return shape_[i]; }
+    auto shape(ssize_t i) const { return shape_[i]; }
 
     T *data() { return buffer_; }
+    const T *data() const { return buffer_; }
     void print_all() const;
 
   private:
     T *buffer_{nullptr};
-    std::array<int64_t, Ndim> strides_{};
+    std::array<ssize_t, Ndim> strides_{};
-    std::array<int64_t, Ndim> shape_{};
+    std::array<ssize_t, Ndim> shape_{};
     uint64_t size_{};
 
-    template <class BinaryOperation> NDView &elemenwise(T val, BinaryOperation op) {
+    template <class BinaryOperation>
+    NDView &elemenwise(T val, BinaryOperation op) {
         for (uint64_t i = 0; i != size_; ++i) {
             buffer_[i] = op(buffer_[i], val);
         }
         return *this;
     }
-    template <class BinaryOperation> NDView &elemenwise(const NDView &other, BinaryOperation op) {
+    template <class BinaryOperation>
+    NDView &elemenwise(const NDView &other, BinaryOperation op) {
         for (uint64_t i = 0; i != size_; ++i) {
             buffer_[i] = op(buffer_[i], other.buffer_[i]);
         }
         return *this;
     }
 };
-template <typename T, int64_t Ndim> void NDView<T, Ndim>::print_all() const {
+
+template <typename T, ssize_t Ndim> void NDView<T, Ndim>::print_all() const {
     for (auto row = 0; row < shape_[0]; ++row) {
         for (auto col = 0; col < shape_[1]; ++col) {
             std::cout << std::setw(3);
@@ -170,9 +189,8 @@ template <typename T, int64_t Ndim> void NDView<T, Ndim>::print_all() const {
     }
 }
 
-
+template <typename T, ssize_t Ndim>
-template <typename T, int64_t Ndim>
+std::ostream &operator<<(std::ostream &os, const NDView<T, Ndim> &arr) {
-std::ostream& operator <<(std::ostream& os, const NDView<T, Ndim>& arr){
     for (auto row = 0; row < arr.shape(0); ++row) {
         for (auto col = 0; col < arr.shape(1); ++col) {
             os << std::setw(3);
@@ -183,5 +201,8 @@ std::ostream& operator <<(std::ostream& os, const NDView<T, Ndim>& arr){
     return os;
 }
 
+template <typename T> NDView<T, 1> make_view(std::vector<T> &vec) {
+    return NDView<T, 1>(vec.data(), {static_cast<ssize_t>(vec.size())});
+}
 
 } // namespace aare

include/aare/NumpyFile.hpp

@@ -1,9 +1,8 @@
 #pragma once
 #include "aare/Dtype.hpp"
-#include "aare/defs.hpp"
 #include "aare/FileInterface.hpp"
 #include "aare/NumpyHelpers.hpp"
-
+#include "aare/defs.hpp"
 
 #include <filesystem>
 #include <iostream>
@@ -11,13 +10,12 @@
 
 namespace aare {
 
-
-
 /**
  * @brief NumpyFile class to read and write numpy files
  * @note derived from FileInterface
  * @note implements all the pure virtual functions from FileInterface
- * @note documentation for the functions can also be found in the FileInterface class
+ * @note documentation for the functions can also be found in the FileInterface
+ * class
  */
 class NumpyFile : public FileInterface {
 
@@ -28,26 +26,35 @@ class NumpyFile : public FileInterface {
      * @param mode file mode (r, w)
      * @param cfg file configuration
      */
-    explicit NumpyFile(const std::filesystem::path &fname, const std::string &mode = "r", FileConfig cfg = {});
+    explicit NumpyFile(const std::filesystem::path &fname,
+                       const std::string &mode = "r", FileConfig cfg = {});
 
     void write(Frame &frame);
     Frame read_frame() override { return get_frame(this->current_frame++); }
-    Frame read_frame(size_t frame_number) override { return get_frame(frame_number); }
+    Frame read_frame(size_t frame_number) override {
+        return get_frame(frame_number);
+    }
 
     std::vector<Frame> read_n(size_t n_frames) override;
-    void read_into(std::byte *image_buf) override { return get_frame_into(this->current_frame++, image_buf); }
+    void read_into(std::byte *image_buf) override {
+        return get_frame_into(this->current_frame++, image_buf);
+    }
     void read_into(std::byte *image_buf, size_t n_frames) override;
     size_t frame_number(size_t frame_index) override { return frame_index; };
     size_t bytes_per_frame() override;
     size_t pixels_per_frame() override;
-    void seek(size_t frame_number) override { this->current_frame = frame_number; }
+    void seek(size_t frame_number) override {
+        this->current_frame = frame_number;
+    }
     size_t tell() override { return this->current_frame; }
     size_t total_frames() const override { return m_header.shape[0]; }
     size_t rows() const override { return m_header.shape[1]; }
     size_t cols() const override { return m_header.shape[2]; }
     size_t bitdepth() const override { return m_header.dtype.bitdepth(); }
 
-    DetectorType detector_type() const override { return DetectorType::Unknown; }
+    DetectorType detector_type() const override {
+        return DetectorType::Unknown;
+    }
 
     /**
      * @brief get the data type of the numpy file
@@ -69,8 +76,9 @@ class NumpyFile : public FileInterface {
      */
     template <typename T, size_t NDim> NDArray<T, NDim> load() {
         NDArray<T, NDim> arr(make_shape<NDim>(m_header.shape));
-        if (fseek(fp, static_cast<int64_t>(header_size), SEEK_SET)) {
+        if (fseek(fp, static_cast<long>(header_size), SEEK_SET)) {
-            throw std::runtime_error(LOCATION + "Error seeking to the start of the data");
+            throw std::runtime_error(LOCATION +
+                                     "Error seeking to the start of the data");
         }
         size_t rc = fread(arr.data(), sizeof(T), arr.size(), fp);
         if (rc != static_cast<size_t>(arr.size())) {
@@ -78,16 +86,20 @@ class NumpyFile : public FileInterface {
         }
         return arr;
     }
-    template <typename A, typename TYPENAME, A Ndim> void write(NDView<TYPENAME, Ndim> &frame) {
+    template <typename A, typename TYPENAME, A Ndim>
+    void write(NDView<TYPENAME, Ndim> &frame) {
         write_impl(frame.data(), frame.total_bytes());
     }
-    template <typename A, typename TYPENAME, A Ndim> void write(NDArray<TYPENAME, Ndim> &frame) {
+    template <typename A, typename TYPENAME, A Ndim>
+    void write(NDArray<TYPENAME, Ndim> &frame) {
         write_impl(frame.data(), frame.total_bytes());
     }
-    template <typename A, typename TYPENAME, A Ndim> void write(NDView<TYPENAME, Ndim> &&frame) {
+    template <typename A, typename TYPENAME, A Ndim>
+    void write(NDView<TYPENAME, Ndim> &&frame) {
         write_impl(frame.data(), frame.total_bytes());
     }
-    template <typename A, typename TYPENAME, A Ndim> void write(NDArray<TYPENAME, Ndim> &&frame) {
+    template <typename A, typename TYPENAME, A Ndim>
+    void write(NDArray<TYPENAME, Ndim> &&frame) {
         write_impl(frame.data(), frame.total_bytes());
     }
 

include/aare/NumpyHelpers.hpp

@@ -40,15 +40,18 @@ bool parse_bool(const std::string &in);
 
 std::string get_value_from_map(const std::string &mapstr);
 
-std::unordered_map<std::string, std::string> parse_dict(std::string in, const std::vector<std::string> &keys);
+std::unordered_map<std::string, std::string>
+parse_dict(std::string in, const std::vector<std::string> &keys);
 
-template <typename T, size_t N> bool in_array(T val, const std::array<T, N> &arr) {
+template <typename T, size_t N>
+bool in_array(T val, const std::array<T, N> &arr) {
     return std::find(std::begin(arr), std::end(arr), val) != std::end(arr);
 }
 bool is_digits(const std::string &str);
 
 aare::Dtype parse_descr(std::string typestring);
-size_t write_header(const std::filesystem::path &fname, const NumpyHeader &header);
+size_t write_header(const std::filesystem::path &fname,
+                    const NumpyHeader &header);
 size_t write_header(std::ostream &out, const NumpyHeader &header);
 
 } // namespace NumpyHelpers

include/aare/Pedestal.hpp

@@ -19,13 +19,13 @@ template <typename SUM_TYPE = double> class Pedestal {
     uint32_t m_samples;
     NDArray<uint32_t, 2> m_cur_samples;
 
-    //TODO! in case of int needs to be changed to uint64_t
+    // TODO! in case of int needs to be changed to uint64_t
     NDArray<SUM_TYPE, 2> m_sum;
     NDArray<SUM_TYPE, 2> m_sum2;
 
-    //Cache mean since it is used over and over in the ClusterFinder
+    // Cache mean since it is used over and over in the ClusterFinder
-    //This optimization is related to the access pattern of the ClusterFinder
+    // This optimization is related to the access pattern of the ClusterFinder
-    //Relies on having more reads than pushes to the pedestal
+    // Relies on having more reads than pushes to the pedestal
     NDArray<SUM_TYPE, 2> m_mean;
 
   public:
@@ -42,9 +42,7 @@ template <typename SUM_TYPE = double> class Pedestal {
     }
     ~Pedestal() = default;
 
-    NDArray<SUM_TYPE, 2> mean() {
+    NDArray<SUM_TYPE, 2> mean() { return m_mean; }
-        return m_mean;
-    }
 
     SUM_TYPE mean(const uint32_t row, const uint32_t col) const {
         return m_mean(row, col);
@@ -71,8 +69,6 @@ template <typename SUM_TYPE = double> class Pedestal {
         return variance_array;
     }
 
-    
-
     NDArray<SUM_TYPE, 2> std() {
         NDArray<SUM_TYPE, 2> standard_deviation_array({m_rows, m_cols});
         for (uint32_t i = 0; i < m_rows * m_cols; i++) {
@@ -83,8 +79,6 @@ template <typename SUM_TYPE = double> class Pedestal {
         return standard_deviation_array;
     }
 
-    
-
     void clear() {
         m_sum = 0;
         m_sum2 = 0;
@@ -92,8 +86,6 @@ template <typename SUM_TYPE = double> class Pedestal {
         m_mean = 0;
     }
 
-    
-
     void clear(const uint32_t row, const uint32_t col) {
         m_sum(row, col) = 0;
         m_sum2(row, col) = 0;
@@ -101,13 +93,11 @@ template <typename SUM_TYPE = double> class Pedestal {
         m_mean(row, col) = 0;
     }
 
-
-
     template <typename T> void push(NDView<T, 2> frame) {
         assert(frame.size() == m_rows * m_cols);
 
         // TODO! move away from m_rows, m_cols
-        if (frame.shape() != std::array<int64_t, 2>{m_rows, m_cols}) {
+        if (frame.shape() != std::array<ssize_t, 2>{m_rows, m_cols}) {
             throw std::runtime_error(
                 "Frame shape does not match pedestal shape");
         }
@@ -128,7 +118,7 @@ template <typename SUM_TYPE = double> class Pedestal {
         assert(frame.size() == m_rows * m_cols);
 
         // TODO! move away from m_rows, m_cols
-        if (frame.shape() != std::array<int64_t, 2>{m_rows, m_cols}) {
+        if (frame.shape() != std::array<ssize_t, 2>{m_rows, m_cols}) {
             throw std::runtime_error(
                 "Frame shape does not match pedestal shape");
         }
@@ -140,9 +130,6 @@ template <typename SUM_TYPE = double> class Pedestal {
         }
     }
 
-
-
-
     template <typename T> void push(Frame &frame) {
         assert(frame.rows() == static_cast<size_t>(m_rows) &&
                frame.cols() == static_cast<size_t>(m_cols));
@@ -170,7 +157,8 @@ template <typename SUM_TYPE = double> class Pedestal {
             m_sum(row, col) += val - m_sum(row, col) / m_samples;
             m_sum2(row, col) += val * val - m_sum2(row, col) / m_samples;
         }
-        //Since we just did a push we know that m_cur_samples(row, col) is at least 1
+        // Since we just did a push we know that m_cur_samples(row, col) is at
+        // least 1
         m_mean(row, col) = m_sum(row, col) / m_cur_samples(row, col);
     }
 
@@ -183,7 +171,8 @@ template <typename SUM_TYPE = double> class Pedestal {
             m_cur_samples(row, col)++;
         } else {
             m_sum(row, col) += val - m_sum(row, col) / m_cur_samples(row, col);
-            m_sum2(row, col) += val * val - m_sum2(row, col) / m_cur_samples(row, col);
+            m_sum2(row, col) +=
+                val * val - m_sum2(row, col) / m_cur_samples(row, col);
         }
     }
 
@@ -191,19 +180,16 @@ template <typename SUM_TYPE = double> class Pedestal {
      * @brief Update the mean of the pedestal. This is used after having done
      * push_no_update. It is not necessary to call this function after push.
      */
-    void update_mean(){
+    void update_mean() { m_mean = m_sum / m_cur_samples; }
-        m_mean = m_sum / m_cur_samples;
-    }
 
-    template<typename T>
+    template <typename T>
-    void push_fast(const uint32_t row, const uint32_t col, const T val_){
+    void push_fast(const uint32_t row, const uint32_t col, const T val_) {
-        //Assume we reached the steady state where all pixels have
+        // Assume we reached the steady state where all pixels have
-        //m_samples samples
+        // m_samples samples
         SUM_TYPE val = static_cast<SUM_TYPE>(val_);
         m_sum(row, col) += val - m_sum(row, col) / m_samples;
         m_sum2(row, col) += val * val - m_sum2(row, col) / m_samples;
         m_mean(row, col) = m_sum(row, col) / m_samples;
     }
-
 };
 } // namespace aare

include/aare/PixelMap.hpp

@@ -1,7 +1,7 @@
 #pragma once
 
-#include "aare/defs.hpp"
 #include "aare/NDArray.hpp"
+#include "aare/defs.hpp"
 
 namespace aare {
 
@@ -10,11 +10,11 @@ NDArray<ssize_t, 2> GenerateMoench05PixelMap();
 NDArray<ssize_t, 2> GenerateMoench05PixelMap1g();
 NDArray<ssize_t, 2> GenerateMoench05PixelMapOld();
 
-//Matterhorn02
+// Matterhorn02
-NDArray<ssize_t, 2>GenerateMH02SingleCounterPixelMap();
+NDArray<ssize_t, 2> GenerateMH02SingleCounterPixelMap();
 NDArray<ssize_t, 3> GenerateMH02FourCounterPixelMap();
 
-//Eiger 
+// Eiger
-NDArray<ssize_t, 2>GenerateEigerFlipRowsPixelMap();
+NDArray<ssize_t, 2> GenerateEigerFlipRowsPixelMap();
 
 } // namespace aare

include/aare/ProducerConsumerQueue.hpp

@@ -18,9 +18,9 @@
 // @author Jordan DeLong (delong.j@fb.com)
 
 // Changes made by PSD Detector Group:
-// Copied: Line 34 constexpr std::size_t hardware_destructive_interference_size = 128; from folly/lang/Align.h
+// Copied: Line 34 constexpr std::size_t hardware_destructive_interference_size
-// Changed extension to .hpp
+// = 128; from folly/lang/Align.h Changed extension to .hpp Changed namespace to
-// Changed namespace to aare
+// aare
 
 #pragma once
 
@@ -45,15 +45,14 @@ template <class T> struct ProducerConsumerQueue {
     ProducerConsumerQueue(const ProducerConsumerQueue &) = delete;
     ProducerConsumerQueue &operator=(const ProducerConsumerQueue &) = delete;
 
-    
+    ProducerConsumerQueue(ProducerConsumerQueue &&other) {
-    ProducerConsumerQueue(ProducerConsumerQueue &&other){
         size_ = other.size_;
         records_ = other.records_;
         other.records_ = nullptr;
         readIndex_ = other.readIndex_.load(std::memory_order_acquire);
         writeIndex_ = other.writeIndex_.load(std::memory_order_acquire);
     }
-    ProducerConsumerQueue &operator=(ProducerConsumerQueue &&other){
+    ProducerConsumerQueue &operator=(ProducerConsumerQueue &&other) {
         size_ = other.size_;
         records_ = other.records_;
         other.records_ = nullptr;
@@ -62,15 +61,16 @@ template <class T> struct ProducerConsumerQueue {
         return *this;
     }
 
-    
+    ProducerConsumerQueue() : ProducerConsumerQueue(2){};
-    ProducerConsumerQueue():ProducerConsumerQueue(2){};
     // size must be >= 2.
     //
     // Also, note that the number of usable slots in the queue at any
     // given time is actually (size-1), so if you start with an empty queue,
     // isFull() will return true after size-1 insertions.
     explicit ProducerConsumerQueue(uint32_t size)
-        : size_(size), records_(static_cast<T *>(std::malloc(sizeof(T) * size))), readIndex_(0), writeIndex_(0) {
+        : size_(size),
+          records_(static_cast<T *>(std::malloc(sizeof(T) * size))),
+          readIndex_(0), writeIndex_(0) {
         assert(size >= 2);
         if (!records_) {
             throw std::bad_alloc();
@@ -154,7 +154,8 @@ template <class T> struct ProducerConsumerQueue {
     }
 
     bool isEmpty() const {
-        return readIndex_.load(std::memory_order_acquire) == writeIndex_.load(std::memory_order_acquire);
+        return readIndex_.load(std::memory_order_acquire) ==
+               writeIndex_.load(std::memory_order_acquire);
     }
 
     bool isFull() const {
@@ -175,7 +176,8 @@ template <class T> struct ProducerConsumerQueue {
     //   be removing items concurrently).
     // * It is undefined to call this from any other thread.
     size_t sizeGuess() const {
-        int ret = writeIndex_.load(std::memory_order_acquire) - readIndex_.load(std::memory_order_acquire);
+        int ret = writeIndex_.load(std::memory_order_acquire) -
+                  readIndex_.load(std::memory_order_acquire);
         if (ret < 0) {
             ret += size_;
         }
@@ -192,7 +194,7 @@ template <class T> struct ProducerConsumerQueue {
     // const uint32_t size_;
     uint32_t size_;
     // T *const records_;
-    T* records_;
+    T *records_;
 
     alignas(hardware_destructive_interference_size) AtomicIndex readIndex_;
     alignas(hardware_destructive_interference_size) AtomicIndex writeIndex_;

include/aare/RawFile.hpp

@@ -1,28 +1,19 @@
 #pragma once
+#include "aare/DetectorGeometry.hpp"
 #include "aare/FileInterface.hpp"
-#include "aare/RawMasterFile.hpp"
 #include "aare/Frame.hpp"
 #include "aare/NDArray.hpp" //for pixel map
+#include "aare/RawMasterFile.hpp"
 #include "aare/RawSubFile.hpp"
 
+#ifdef AARE_TESTS
+#include "../tests/friend_test.hpp"
+#endif
 
 #include <optional>
 
 namespace aare {
 
-struct ModuleConfig {
-    int module_gap_row{};
-    int module_gap_col{};
-
-    bool operator==(const ModuleConfig &other) const {
-        if (module_gap_col != other.module_gap_col)
-            return false;
-        if (module_gap_row != other.module_gap_row)
-            return false;
-        return true;
-    }
-};
-
 /**
  * @brief Class to read .raw files. The class will parse the master file
  * to find the correct geometry for the frames.
@@ -30,22 +21,12 @@ struct ModuleConfig {
  * Consider using that unless you need raw file specific functionality.
  */
 class RawFile : public FileInterface {
-    size_t n_subfiles{}; //f0,f1...fn
-    size_t n_subfile_parts{}; // d0,d1...dn
-    //TODO! move to vector of SubFile instead of pointers
-    std::vector<std::vector<RawSubFile *>> subfiles; //subfiles[f0,f1...fn][d0,d1...dn]
-    // std::vector<xy> positions;
 
-    ModuleConfig cfg{0, 0};
+    std::vector<std::unique_ptr<RawSubFile>> m_subfiles;
 
     RawMasterFile m_master;
-
     size_t m_current_frame{};
 
-    // std::vector<ModuleGeometry> m_module_pixel_0;
-    // size_t m_rows{};
-    // size_t m_cols{};
-
     DetectorGeometry m_geometry;
 
   public:
@@ -56,7 +37,7 @@ class RawFile : public FileInterface {
 
      */
     RawFile(const std::filesystem::path &fname, const std::string &mode = "r");
-    virtual ~RawFile() override;
+    virtual ~RawFile() override = default;
 
     Frame read_frame() override;
     Frame read_frame(size_t frame_number) override;
@@ -64,10 +45,10 @@ class RawFile : public FileInterface {
     void read_into(std::byte *image_buf) override;
     void read_into(std::byte *image_buf, size_t n_frames) override;
 
-    //TODO! do we need to adapt the API? 
+    // TODO! do we need to adapt the API?
     void read_into(std::byte *image_buf, DetectorHeader *header);
-    void read_into(std::byte *image_buf, size_t n_frames, DetectorHeader *header);
+    void read_into(std::byte *image_buf, size_t n_frames,
-
+                   DetectorHeader *header);
 
     size_t frame_number(size_t frame_index) override;
     size_t bytes_per_frame() override;
@@ -79,35 +60,14 @@ class RawFile : public FileInterface {
     size_t rows() const override;
     size_t cols() const override;
     size_t bitdepth() const override;
-    xy geometry();
+    size_t n_modules() const;
-    size_t n_mod() const;
+    size_t n_modules_in_roi() const;
+    xy geometry() const;
 
     RawMasterFile master() const;
 
-
-
     DetectorType detector_type() const override;
 
-  private:
-
-    /**
-     * @brief read the frame at the given frame index into the image buffer
-     * @param frame_number frame number to read
-     * @param image_buf buffer to store the frame
-     */
-
-    void get_frame_into(size_t frame_index, std::byte *frame_buffer, DetectorHeader *header = nullptr);
-
-
-    /**
-     * @brief get the frame at the given frame index
-     * @param frame_number frame number to read
-     * @return Frame
-     */
-    Frame get_frame(size_t frame_index);
-
-
-
     /**
      * @brief read the header of the file
      * @param fname path to the data subfile
@@ -115,12 +75,24 @@ class RawFile : public FileInterface {
      */
     static DetectorHeader read_header(const std::filesystem::path &fname);
 
-    // void update_geometry_with_roi();
+  private:
-    int find_number_of_subfiles();
+    /**
+     * @brief read the frame at the given frame index into the image buffer
+     * @param frame_number frame number to read
+     * @param image_buf buffer to store the frame
+     */
+
+    void get_frame_into(size_t frame_index, std::byte *frame_buffer,
+                        DetectorHeader *header = nullptr);
+
+    /**
+     * @brief get the frame at the given frame index
+     * @param frame_number frame number to read
+     * @return Frame
+     */
+    Frame get_frame(size_t frame_index);
 
     void open_subfiles();
-    void find_geometry();
 };
 
-
 } // namespace aare

include/aare/RawMasterFile.hpp

@@ -1,5 +1,6 @@
 #pragma once
 #include "aare/defs.hpp"
+#include <algorithm>
 #include <filesystem>
 #include <fmt/format.h>
 #include <fstream>
@@ -45,7 +46,7 @@ class ScanParameters {
     int m_start = 0;
     int m_stop = 0;
     int m_step = 0;
-    //TODO! add settleTime, requires string to time conversion
+    // TODO! add settleTime, requires string to time conversion
 
   public:
     ScanParameters(const std::string &par);
@@ -61,7 +62,6 @@ class ScanParameters {
     void increment_stop();
 };
 
-
 /**
  * @brief Class for parsing a master file either in our .json format or the old
  * .raw format
@@ -78,8 +78,10 @@ class RawMasterFile {
     size_t m_pixels_y{};
     size_t m_pixels_x{};
     size_t m_bitdepth{};
+    uint8_t m_quad = 0;
 
     xy m_geometry{};
+    xy m_udp_interfaces_per_module{1, 1};
 
     size_t m_max_frames_per_file{};
     // uint32_t m_adc_mask{}; // TODO! implement reading
@@ -97,11 +99,9 @@ class RawMasterFile {
     std::optional<size_t> m_digital_samples;
     std::optional<size_t> m_transceiver_samples;
     std::optional<size_t> m_number_of_rows;
-    std::optional<uint8_t> m_quad;
 
     std::optional<ROI> m_roi;
 
-
   public:
     RawMasterFile(const std::filesystem::path &fpath);
 
@@ -117,26 +117,27 @@ class RawMasterFile {
     size_t max_frames_per_file() const;
     size_t bitdepth() const;
     size_t frame_padding() const;
+    xy udp_interfaces_per_module() const;
     const FrameDiscardPolicy &frame_discard_policy() const;
 
     size_t total_frames_expected() const;
     xy geometry() const;
+    size_t n_modules() const;
+    uint8_t quad() const;
 
     std::optional<size_t> analog_samples() const;
     std::optional<size_t> digital_samples() const;
     std::optional<size_t> transceiver_samples() const;
     std::optional<size_t> number_of_rows() const;
-    std::optional<uint8_t> quad() const;
-
 
     std::optional<ROI> roi() const;
 
-
     ScanParameters scan_parameters() const;
 
   private:
     void parse_json(const std::filesystem::path &fpath);
     void parse_raw(const std::filesystem::path &fpath);
+    void retrieve_geometry();
 };
 
 } // namespace aare

include/aare/RawSubFile.hpp

@@ -10,23 +10,34 @@
 namespace aare {
 
 /**
- * @brief Class to read a singe subfile written in .raw format. Used from RawFile to read
+ * @brief Class to read a singe subfile written in .raw format. Used from
- * the entire detector. Can be used directly to read part of the image.
+ * RawFile to read the entire detector. Can be used directly to read part of the
+ * image.
  */
 class RawSubFile {
   protected:
     std::ifstream m_file;
     DetectorType m_detector_type;
     size_t m_bitdepth;
-    std::filesystem::path m_fname;
+    std::filesystem::path m_path; //!< path to the subfile
+    std::string m_base_name;      //!< base name used for formatting file names
+    size_t m_offset{}; //!< file index of the first file, allow starting at non
+                       //!< zero file
+    size_t m_total_frames{}; //!< total number of frames in the series of files
     size_t m_rows{};
     size_t m_cols{};
     size_t m_bytes_per_frame{};
-    size_t n_frames{};
+
+    int m_module_index{};
+    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
+
     uint32_t m_pos_row{};
     uint32_t m_pos_col{};
 
-    
     std::optional<NDArray<ssize_t, 2>> m_pixel_map;
 
   public:
@@ -40,12 +51,14 @@ class RawSubFile {
      * @throws std::invalid_argument if the detector,type pair is not supported
      */
     RawSubFile(const std::filesystem::path &fname, DetectorType detector,
-               size_t rows, size_t cols, size_t bitdepth, uint32_t pos_row = 0, uint32_t pos_col = 0);
+               size_t rows, size_t cols, size_t bitdepth, uint32_t pos_row = 0,
+               uint32_t pos_col = 0);
 
     ~RawSubFile() = default;
     /**
      * @brief Seek to the given frame number
-     * @note Puts the file pointer at the start of the header, not the start of the data
+     * @note Puts the file pointer at the start of the header, not the start of
+     * the data
      * @param frame_index frame position in file to seek to
      * @throws std::runtime_error if the frame number is out of range
      */
@@ -53,6 +66,8 @@ 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);
@@ -66,10 +81,15 @@ class RawSubFile {
     size_t pixels_per_frame() const { return m_rows * m_cols; }
     size_t bytes_per_pixel() const { return m_bitdepth / bits_per_byte; }
 
-private:
+    size_t frames_in_file() const { return m_total_frames; }
-  template <typename T>
-  void read_with_map(std::byte *image_buf);
 
+  private:
+    template <typename T> void read_with_map(std::byte *image_buf);
+
+    void parse_fname(const std::filesystem::path &fname);
+    void scan_files();
+    void open_file(size_t file_index);
+    std::filesystem::path fpath(size_t file_index) const;
 };
 
 } // namespace aare

include/aare/VarClusterFinder.hpp

@@ -28,7 +28,7 @@ template <typename T> class VarClusterFinder {
     };
 
   private:
-    const std::array<int64_t, 2> shape_;
+    const std::array<ssize_t, 2> shape_;
     NDView<T, 2> original_;
     NDArray<int, 2> labeled_;
     NDArray<int, 2> peripheral_labeled_;
@@ -38,8 +38,10 @@ template <typename T> class VarClusterFinder {
     bool use_noise_map = false;
     int peripheralThresholdFactor_ = 5;
     int current_label;
-    const std::array<int, 4> di{{0, -1, -1, -1}};              // row ### 8-neighbour by scaning from left to right
+    const std::array<int, 4> di{
-    const std::array<int, 4> dj{{-1, -1, 0, 1}};               // col ### 8-neighbour by scaning from top to bottom
+        {0, -1, -1, -1}}; // row ### 8-neighbour by scaning from left to right
+    const std::array<int, 4> dj{
+        {-1, -1, 0, 1}}; // col ### 8-neighbour by scaning from top to bottom
     const std::array<int, 8> di_{{0, 0, -1, 1, -1, 1, -1, 1}}; // row
     const std::array<int, 8> dj_{{-1, 1, 0, 0, 1, -1, -1, 1}}; // col
     std::map<int, int> child; // heirachy: key: child; val: parent
@@ -50,7 +52,8 @@ template <typename T> class VarClusterFinder {
 
   public:
     VarClusterFinder(Shape<2> shape, T threshold)
-        : shape_(shape), labeled_(shape, 0), peripheral_labeled_(shape, 0), binary_(shape), threshold_(threshold) {
+        : shape_(shape), labeled_(shape, 0), peripheral_labeled_(shape, 0),
+          binary_(shape), threshold_(threshold) {
         hits.reserve(2000);
     }
 
@@ -60,7 +63,9 @@ template <typename T> class VarClusterFinder {
         noiseMap = noise_map;
         use_noise_map = true;
     }
-    void set_peripheralThresholdFactor(int factor) { peripheralThresholdFactor_ = factor; }
+    void set_peripheralThresholdFactor(int factor) {
+        peripheralThresholdFactor_ = factor;
+    }
     void find_clusters(NDView<T, 2> img);
     void find_clusters_X(NDView<T, 2> img);
     void rec_FillHit(int clusterIndex, int i, int j);
@@ -144,7 +149,8 @@ template <typename T> int VarClusterFinder<T>::check_neighbours(int i, int j) {
     }
 }
 
-template <typename T> void VarClusterFinder<T>::find_clusters(NDView<T, 2> img) {
+template <typename T>
+void VarClusterFinder<T>::find_clusters(NDView<T, 2> img) {
     original_ = img;
     labeled_ = 0;
     peripheral_labeled_ = 0;
@@ -156,7 +162,8 @@ template <typename T> void VarClusterFinder<T>::find_clusters(NDView<T, 2> img)
     store_clusters();
 }
 
-template <typename T> void VarClusterFinder<T>::find_clusters_X(NDView<T, 2> img) {
+template <typename T>
+void VarClusterFinder<T>::find_clusters_X(NDView<T, 2> img) {
     original_ = img;
     int clusterIndex = 0;
     for (int i = 0; i < shape_[0]; ++i) {
@@ -175,7 +182,8 @@ template <typename T> void VarClusterFinder<T>::find_clusters_X(NDView<T, 2> img
     h_size.clear();
 }
 
-template <typename T> void VarClusterFinder<T>::rec_FillHit(int clusterIndex, int i, int j) {
+template <typename T>
+void VarClusterFinder<T>::rec_FillHit(int clusterIndex, int i, int j) {
     // printf("original_(%d, %d)=%f\n", i, j, original_(i,j));
     // printf("h_size[%d].size=%d\n", clusterIndex, h_size[clusterIndex].size);
     if (h_size[clusterIndex].size < MAX_CLUSTER_SIZE) {
@@ -203,11 +211,15 @@ template <typename T> void VarClusterFinder<T>::rec_FillHit(int clusterIndex, in
             } else {
                 // if (h_size[clusterIndex].size < MAX_CLUSTER_SIZE){
                 //     h_size[clusterIndex].size += 1;
-                //     h_size[clusterIndex].rows[h_size[clusterIndex].size] = row;
+                //     h_size[clusterIndex].rows[h_size[clusterIndex].size] =
-                //     h_size[clusterIndex].cols[h_size[clusterIndex].size] = col;
+                //     row; h_size[clusterIndex].cols[h_size[clusterIndex].size]
-                //     h_size[clusterIndex].enes[h_size[clusterIndex].size] = original_(row, col);
+                //     = col;
+                //     h_size[clusterIndex].enes[h_size[clusterIndex].size] =
+                //     original_(row, col);
                 // }// ? weather to include peripheral pixels
-                original_(row, col) = 0; // remove peripheral pixels, to avoid potential influence for pedestal updating
+                original_(row, col) =
+                    0; // remove peripheral pixels, to avoid potential influence
+                       // for pedestal updating
             }
         }
     }
@@ -275,8 +287,8 @@ template <typename T> void VarClusterFinder<T>::store_clusters() {
     for (int i = 0; i < shape_[0]; ++i) {
         for (int j = 0; j < shape_[1]; ++j) {
             if (labeled_(i, j) != 0 || false
-                // (i-1 >= 0 and labeled_(i-1, j) != 0) or // another circle of peripheral pixels
+                // (i-1 >= 0 and labeled_(i-1, j) != 0) or // another circle of
-                // (j-1 >= 0 and labeled_(i, j-1) != 0) or
+                // peripheral pixels (j-1 >= 0 and labeled_(i, j-1) != 0) or
                 // (i+1 < shape_[0] and labeled_(i+1, j) != 0) or
                 // (j+1 < shape_[1] and labeled_(i, j+1) != 0)
             ) {

include/aare/algorithm.hpp

@@ -1,9 +1,9 @@
 
 #pragma once
+#include <aare/NDArray.hpp>
 #include <algorithm>
 #include <array>
 #include <vector>
-#include <aare/NDArray.hpp>
 
 namespace aare {
 /**
@@ -18,23 +18,21 @@ namespace aare {
  *
  */
 template <typename T>
-size_t last_smaller(const T* first, const T* last, T val) {
+size_t last_smaller(const T *first, const T *last, T val) {
-    for (auto iter = first+1; iter != last; ++iter) {
+    for (auto iter = first + 1; iter != last; ++iter) {
         if (*iter >= val) {
-            return std::distance(first, iter-1);
+            return std::distance(first, iter - 1);
         }
     }
-    return std::distance(first, last-1);
+    return std::distance(first, last - 1);
 }
 
-template <typename T>
+template <typename T> size_t last_smaller(const NDArray<T, 1> &arr, T val) {
-size_t last_smaller(const NDArray<T, 1>& arr, T val) {
     return last_smaller(arr.begin(), arr.end(), val);
 }
 
-template <typename T>
+template <typename T> size_t last_smaller(const std::vector<T> &vec, T val) {
-size_t last_smaller(const std::vector<T>& vec, T val) {
+    return last_smaller(vec.data(), vec.data() + vec.size(), val);
-    return last_smaller(vec.data(), vec.data()+vec.size(), val);
 }
 
 /**
@@ -48,64 +46,67 @@ size_t last_smaller(const std::vector<T>& vec, T val) {
  * @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) {
+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);
+    return std::distance(first, last - 1);
 }
 
-template <typename T>
+template <typename T> size_t first_larger(const NDArray<T, 1> &arr, T val) {
-size_t first_larger(const NDArray<T, 1>& arr, T val) {
     return first_larger(arr.begin(), arr.end(), val);
 }
 
-template <typename T>
+template <typename T> size_t first_larger(const std::vector<T> &vec, T val) {
-size_t first_larger(const std::vector<T>& vec, T val) {
+    return first_larger(vec.data(), vec.data() + vec.size(), 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.
+ * 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) {
+size_t nearest_index(const T *first, const T *last, T val) {
-    auto iter = std::min_element(first, last,
+    auto iter = std::min_element(first, last, [val](T a, T b) {
-    [val](T a, T b) {
         return std::abs(a - val) < std::abs(b - val);
     });
     return std::distance(first, iter);
 }
 
-template <typename T>
+template <typename T> size_t nearest_index(const NDArray<T, 1> &arr, T val) {
-size_t nearest_index(const NDArray<T, 1>& arr, T val) {
     return nearest_index(arr.begin(), arr.end(), val);
 }
 
-template <typename T>
+template <typename T> size_t nearest_index(const std::vector<T> &vec, T val) {
-size_t nearest_index(const std::vector<T>& vec, T val) {
+    return nearest_index(vec.data(), vec.data() + vec.size(), 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) {
+size_t nearest_index(const std::array<T, N> &arr, T val) {
-    return nearest_index(arr.data(), arr.data()+arr.size(), val);
+    return nearest_index(arr.data(), arr.data() + arr.size(), val);
 }
 
-template <typename T>
+template <typename T> std::vector<T> cumsum(const std::vector<T> &vec) {
-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;
 }
 
-
+template <typename Container> bool all_equal(const Container &c) {
+    if (!c.empty() &&
+        std::all_of(begin(c), end(c),
+                    [c](const typename Container::value_type &element) {
+                        return element == c.front();
+                    }))
+        return true;
+    return false;
+}
 
 } // namespace aare

include/aare/decode.hpp

@@ -1,13 +1,27 @@
 #pragma once
 
-#include <cstdint>
 #include <aare/NDView.hpp>
+#include <cstdint>
+#include <vector>
 namespace aare {
 
-
 uint16_t adc_sar_05_decode64to16(uint64_t input);
 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_05_decode64to16(NDView<uint64_t, 2> input,
-void adc_sar_04_decode64to16(NDView<uint64_t, 2> input, NDView<uint16_t,2> output);
+                             NDView<uint16_t, 2> output);
+void adc_sar_04_decode64to16(NDView<uint64_t, 2> input,
+                             NDView<uint16_t, 2> output);
+
+/**
+ * @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

@@ -3,16 +3,15 @@
 #include "aare/Dtype.hpp"
 
 #include <array>
-#include <stdexcept>
 #include <cassert>
 #include <cstdint>
 #include <cstring>
+#include <stdexcept>
 #include <string>
 #include <string_view>
 #include <variant>
 #include <vector>
 
-
 /**
  * @brief LOCATION macro to get the current location in the code
  */
@@ -20,28 +19,24 @@
     std::string(__FILE__) + std::string(":") + std::to_string(__LINE__) +      \
         ":" + std::string(__func__) + ":"
 
-
-
 #ifdef AARE_CUSTOM_ASSERT
-#define AARE_ASSERT(expr)\
+#define AARE_ASSERT(expr)                                                      \
-    if (expr)\
+    if (expr) {                                                                \
-        {}\
+    } else                                                                     \
-    else\
         aare::assert_failed(LOCATION + " Assertion failed: " + #expr + "\n");
 #else
-#define AARE_ASSERT(cond)\
+#define AARE_ASSERT(cond)                                                      \
-    do { (void)sizeof(cond); } while(0)
+    do {                                                                       \
+        (void)sizeof(cond);                                                    \
+    } while (0)
 #endif
 
-
 namespace aare {
 
 inline constexpr size_t bits_per_byte = 8;
 
 void assert_failed(const std::string &msg);
 
-
-
 class DynamicCluster {
   public:
     int cluster_sizeX;
@@ -179,50 +174,22 @@ template <typename T> struct t_xy {
 };
 using xy = t_xy<uint32_t>;
 
+struct ROI {
+    ssize_t xmin{};
+    ssize_t xmax{};
+    ssize_t ymin{};
+    ssize_t ymax{};
 
-/**
+    ssize_t height() const { return ymax - ymin; }
- * @brief Class to hold the geometry of a module. Where pixel 0 is located and the size of the module
+    ssize_t width() const { return xmax - xmin; }
- */
+    bool contains(ssize_t x, ssize_t y) const {
-struct ModuleGeometry{
-    int origin_x{};
-    int origin_y{};
-    int height{};
-    int width{};
-    int row_index{};
-    int col_index{}; 
-};
-
-/**
- * @brief Class to hold the geometry of a detector. Number of modules, their size and where pixel 0 
- * for each module is located
- */
-struct DetectorGeometry{
-    int modules_x{};
-    int modules_y{};
-    int pixels_x{};
-    int pixels_y{};
-    int module_gap_row{};
-    int module_gap_col{};
-    std::vector<ModuleGeometry> module_pixel_0;
-};
-
-struct ROI{
-    int64_t xmin{};
-    int64_t xmax{};
-    int64_t ymin{};
-    int64_t ymax{};
-  
-    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;
     }
-  };
+};
 
+using dynamic_shape = std::vector<ssize_t>;
 
-using dynamic_shape = std::vector<int64_t>;
+// TODO! Can we uniform enums between the libraries?
-
-//TODO! Can we uniform enums between the libraries?
 
 /**
  * @brief Enum class to identify different detectors.
@@ -230,7 +197,7 @@ using dynamic_shape = std::vector<int64_t>;
  * Different spelling to avoid confusion with the slsDetectorPackage
  */
 enum class DetectorType {
-    //Standard detectors match the enum values from slsDetectorPackage
+    // Standard detectors match the enum values from slsDetectorPackage
     Generic,
     Eiger,
     Gotthard,
@@ -241,8 +208,9 @@ enum class DetectorType {
     Gotthard2,
     Xilinx_ChipTestBoard,
 
-    //Additional detectors used for defining processing. Variants of the standard ones.
+    // Additional detectors used for defining processing. Variants of the
-    Moench03=100,
+    // standard ones.
+    Moench03 = 100,
     Moench03_old,
     Unknown
 };

include/aare/geo_helpers.hpp

@@ -1,16 +0,0 @@
-#pragma once
-#include "aare/defs.hpp"
-#include "aare/RawMasterFile.hpp" //ROI refactor away
-namespace aare{
-
-/**
- * @brief Update the detector geometry given a region of interest
- * 
- * @param geo 
- * @param roi 
- * @return DetectorGeometry 
- */
-DetectorGeometry update_geometry_with_roi(DetectorGeometry geo, ROI roi);
-
-
-} // namespace aare

include/aare/logger.hpp

@@ -0,0 +1,141 @@
+#pragma once
+/*Utility to log to console*/
+
+#include <iostream>
+#include <sstream>
+#include <sys/time.h>
+
+namespace aare {
+
+#define RED      "\x1b[31m"
+#define GREEN    "\x1b[32m"
+#define YELLOW   "\x1b[33m"
+#define BLUE     "\x1b[34m"
+#define MAGENTA  "\x1b[35m"
+#define CYAN     "\x1b[36m"
+#define GRAY     "\x1b[37m"
+#define DARKGRAY "\x1b[30m"
+
+#define BG_BLACK   "\x1b[48;5;232m"
+#define BG_RED     "\x1b[41m"
+#define BG_GREEN   "\x1b[42m"
+#define BG_YELLOW  "\x1b[43m"
+#define BG_BLUE    "\x1b[44m"
+#define BG_MAGENTA "\x1b[45m"
+#define BG_CYAN    "\x1b[46m"
+#define RESET      "\x1b[0m"
+#define BOLD       "\x1b[1m"
+
+enum TLogLevel {
+    logERROR,
+    logWARNING,
+    logINFOBLUE,
+    logINFOGREEN,
+    logINFORED,
+    logINFOCYAN,
+    logINFOMAGENTA,
+    logINFO,
+    logDEBUG, // constructors, destructors etc. should still give too much
+              // output
+    logDEBUG1,
+    logDEBUG2,
+    logDEBUG3,
+    logDEBUG4,
+    logDEBUG5
+};
+
+// Compiler should optimize away anything below this value
+#ifndef AARE_LOG_LEVEL
+#define AARE_LOG_LEVEL                                                         \
+    "LOG LEVEL NOT SET IN CMAKE" // This is configured in the main
+                                 // CMakeLists.txt
+#endif
+
+#define __AT__                                                                 \
+    std::string(__FILE__) + std::string("::") + std::string(__func__) +        \
+        std::string("(): ")
+#define __SHORT_FORM_OF_FILE__                                                 \
+    (strrchr(__FILE__, '/') ? strrchr(__FILE__, '/') + 1 : __FILE__)
+#define __SHORT_AT__                                                           \
+    std::string(__SHORT_FORM_OF_FILE__) + std::string("::") +                  \
+        std::string(__func__) + std::string("(): ")
+
+class Logger {
+    std::ostringstream os;
+    TLogLevel m_level = AARE_LOG_LEVEL;
+
+  public:
+    Logger() = default;
+    explicit Logger(TLogLevel level) : m_level(level){};
+    ~Logger() {
+        // output in the destructor to allow for << syntax
+        os << RESET << '\n';
+        std::clog << os.str() << std::flush; // Single write
+    }
+
+    static TLogLevel &
+    ReportingLevel() { // singelton eeh TODO! Do we need a runtime option?
+        static TLogLevel reportingLevel = logDEBUG5;
+        return reportingLevel;
+    }
+
+    // Danger this buffer need as many elements as TLogLevel
+    static const char *Color(TLogLevel level) noexcept {
+        static const char *const colors[] = {
+            RED BOLD, YELLOW BOLD, BLUE,  GREEN, RED,   CYAN,  MAGENTA,
+            RESET,    RESET,       RESET, RESET, RESET, RESET, RESET};
+        // out of bounds
+        if (level < 0 || level >= sizeof(colors) / sizeof(colors[0])) {
+            return RESET;
+        }
+        return colors[level];
+    }
+
+    // Danger this buffer need as many elements as TLogLevel
+    static std::string ToString(TLogLevel level) {
+        static const char *const buffer[] = {
+            "ERROR",  "WARNING", "INFO",   "INFO",  "INFO",
+            "INFO",   "INFO",    "INFO",   "DEBUG", "DEBUG1",
+            "DEBUG2", "DEBUG3",  "DEBUG4", "DEBUG5"};
+        // out of bounds
+        if (level < 0 || level >= sizeof(buffer) / sizeof(buffer[0])) {
+            return "UNKNOWN";
+        }
+        return buffer[level];
+    }
+
+    std::ostringstream &Get() {
+        os << Color(m_level) << "- " << Timestamp() << " " << ToString(m_level)
+           << ": ";
+        return os;
+    }
+
+    static std::string Timestamp() {
+        constexpr size_t buffer_len = 12;
+        char buffer[buffer_len];
+        time_t t;
+        ::time(&t);
+        tm r;
+        strftime(buffer, buffer_len, "%X", localtime_r(&t, &r));
+        buffer[buffer_len - 1] = '\0';
+        struct timeval tv;
+        gettimeofday(&tv, nullptr);
+        constexpr size_t result_len = 100;
+        char result[result_len];
+        snprintf(result, result_len, "%s.%03ld", buffer,
+                 static_cast<long>(tv.tv_usec) / 1000);
+        result[result_len - 1] = '\0';
+        return result;
+    }
+};
+
+// TODO! Do we need to keep the runtime option?
+#define LOG(level)                                                             \
+    if (level > AARE_LOG_LEVEL)                                                \
+        ;                                                                      \
+    else if (level > aare::Logger::ReportingLevel())                           \
+        ;                                                                      \
+    else                                                                       \
+        aare::Logger(level).Get()
+
+} // namespace aare

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

@@ -1,11 +1,11 @@
 #include <thread>
-#include <vector>
 #include <utility>
+#include <vector>
 
 namespace aare {
 
-    template<typename F>
+template <typename F>
-    void RunInParallel(F func, const std::vector<std::pair<int, int>>& tasks) {
+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) {
@@ -14,5 +14,5 @@ namespace aare {
     for (auto &thread : threads) {
         thread.join();
     }
-    }
+}
 } // namespace aare

pyproject.toml

@@ -1,22 +1,40 @@
+[tool.scikit-build.metadata.version]
+provider = "scikit_build_core.metadata.regex"
+input = "VERSION"
+regex = '^(?P<version>\d+(?:\.\d+)*(?:[\.\+\w]+)?)$'
+result = "{version}" 
+
 [build-system]
 requires = ["scikit-build-core>=0.10", "pybind11", "numpy"]
 build-backend = "scikit_build_core.build"
 
 [project]
 name = "aare"
-version = "2025.4.1"
+dynamic = ["version"]
+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"
 
+
 [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()
@@ -62,10 +63,16 @@ 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

@@ -1,14 +1,87 @@
-
+from . import _aare
-from ._aare import ClusterFinder_Cluster3x3i
 import numpy as np
 
+_supported_cluster_sizes = [(2,2), (3,3), (5,5), (7,7), (9,9),]
+
+def _type_to_char(dtype):
+    if dtype == np.int32:
+        return 'i'
+    elif dtype == np.float32:
+        return 'f'
+    elif dtype == np.float64:
+        return 'd'
+    else:
+        raise ValueError(f"Unsupported dtype: {dtype}. Only np.int32, np.float32, and np.float64 are supported.")
+
+def _get_class(name, cluster_size, dtype):
+    """
+    Helper function to get the class based on the name, cluster size, and dtype.
+    """
+    try:
+        class_name = f"{name}_Cluster{cluster_size[0]}x{cluster_size[1]}{_type_to_char(dtype)}"
+        cls = getattr(_aare, class_name)
+    except AttributeError:
+        raise ValueError(f"Unsupported combination of type and cluster size: {dtype}/{cluster_size} when requesting {class_name}")
+    return cls
+
+
+
 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):
+    cls = _get_class("ClusterFinder", cluster_size, dtype)
-        return ClusterFinder_Cluster3x3i(image_size, n_sigma = n_sigma, capacity=capacity)
+    return cls(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++.
+    """
+
+    cls = _get_class("ClusterFinderMT", cluster_size, dtype)
+    return cls(image_size, n_sigma=n_sigma, capacity=capacity, n_threads=n_threads)
+
+
+
+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++.
+    """
+
+    cls = _get_class("ClusterCollector", cluster_size, dtype)
+    return cls(clusterfindermt)
+
+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++.
+    """
+
+    cls = _get_class("ClusterFileSink", clusterfindermt.cluster_size, dtype)
+    return cls(clusterfindermt, cluster_file)
+
+
+def ClusterFile(fname, cluster_size=(3,3), dtype=np.int32, chunk_size = 1000):
+    """
+    Factory function to create a ClusterFile object. Provides a cleaner syntax for
+    the templated ClusterFile in C++.
+
+    .. code-block:: python
+
+        from aare import ClusterFile
+        
+        with ClusterFile("clusters.clust", cluster_size=(3,3), dtype=np.int32) as cf:
+            # cf is now a ClusterFile_Cluster3x3i object but you don't need to know that.
+            for clusters in cf:
+                # Loop over clusters in chunks of 1000 
+                # The type of clusters will be a ClusterVector_Cluster3x3i in this case
+
+    """
+
+    cls = _get_class("ClusterFile", cluster_size, dtype)
+    return cls(fname, chunk_size=chunk_size)

python/aare/__init__.py

@@ -5,18 +5,22 @@ from . import _aare
 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_Cluster3x3i as ClusterFile
 from ._aare import hitmap
 from ._aare import ROI
 
 # from ._aare import ClusterFinderMT, ClusterCollector, ClusterFileSink, ClusterVector_i
 
-from .ClusterFinder import ClusterFinder
+from .ClusterFinder import ClusterFinder, ClusterCollector, ClusterFinderMT, ClusterFileSink, ClusterFile
 from .ClusterVector import ClusterVector
 
-from ._aare import fit_gaus, fit_pol1
+
+from ._aare import fit_gaus, fit_pol1, fit_scurve, fit_scurve2
 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

python/aare/func.py

@@ -1 +1 @@
-from ._aare import gaus, pol1
+from ._aare import gaus, pol1, scurve, scurve2

python/examples/play.py

@@ -1,79 +1,89 @@
 import sys
 sys.path.append('/home/l_msdetect/erik/aare/build')
 
-from aare._aare import ClusterVector_i, Interpolator
 
-import pickle 
+from aare import RawSubFile, DetectorType, RawFile
-import numpy as np
+
-import matplotlib.pyplot as plt
+from pathlib import Path
-import boost_histogram as bh
+path = Path("/home/l_msdetect/erik/data/aare-test-data/raw/jungfrau/")
-import torch
+f = RawSubFile(path/"jungfrau_single_d0_f0_0.raw", DetectorType.Jungfrau, 512, 1024, 16)
-import math
+
-import time
+# f = RawFile(path/"jungfrau_single_master_0.json")
+
+
+# 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
 
 
 
-def gaussian_2d(mx, my, sigma = 1, res=100, grid_size = 2):
+# def gaussian_2d(mx, my, sigma = 1, res=100, grid_size = 2):
-    """
+#     """
-    Generate a 2D gaussian as position mx, my, with sigma=sigma. 
+#     Generate a 2D gaussian as position mx, my, with sigma=sigma. 
-    The gaussian is placed on a 2x2 pixel matrix with resolution 
+#     The gaussian is placed on a 2x2 pixel matrix with resolution 
-    res in one dimesion.
+#     res in one dimesion.
-    """
+#     """
-    x = torch.linspace(0, pixel_size*grid_size, res)
+#     x = torch.linspace(0, pixel_size*grid_size, res)
-    x,y = torch.meshgrid(x,x, indexing="ij")
+#     x,y = torch.meshgrid(x,x, indexing="ij")
-    return 1 / (2*math.pi*sigma**2) * \
+#     return 1 / (2*math.pi*sigma**2) * \
-      torch.exp(-((x - my)**2 / (2*sigma**2) + (y - mx)**2 / (2*sigma**2)))
+#       torch.exp(-((x - my)**2 / (2*sigma**2) + (y - mx)**2 / (2*sigma**2)))
 
-scale = 1000 #Scale factor when converting to integer
+# scale = 1000 #Scale factor when converting to integer
-pixel_size = 25 #um
+# pixel_size = 25 #um
-grid = 2
+# grid = 2
-resolution = 100
+# resolution = 100
-sigma_um = 10
+# sigma_um = 10
-xa = np.linspace(0,grid*pixel_size,resolution)
+# xa = np.linspace(0,grid*pixel_size,resolution)
-ticks = [0, 25, 50]
+# ticks = [0, 25, 50]
 
-hit = np.array((20,20))
+# hit = np.array((20,20))
-etahist_fname = "/home/l_msdetect/erik/tmp/test_hist.pkl"
+# etahist_fname = "/home/l_msdetect/erik/tmp/test_hist.pkl"
 
-local_resolution = 99
+# local_resolution = 99
-grid_size = 3
+# grid_size = 3
-xaxis = np.linspace(0,grid_size*pixel_size, local_resolution)
+# 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)
+# 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 = 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.numpy()
-pixels = (pixels*scale).astype(np.int32)
+# pixels = (pixels*scale).astype(np.int32)
-v = ClusterVector_i(3,3)
+# v = ClusterVector_i(3,3)
-v.push_back(1,1, pixels)
+# v.push_back(1,1, pixels)
 
-with open(etahist_fname, "rb") as f:
+# with open(etahist_fname, "rb") as f:
-        hist = pickle.load(f)
+#         hist = pickle.load(f)
-eta = hist.view().copy()
+# eta = hist.view().copy()
-etabinsx = np.array(hist.axes.edges.T[0].flat)
+# etabinsx = np.array(hist.axes.edges.T[0].flat)
-etabinsy = np.array(hist.axes.edges.T[1].flat)
+# etabinsy = np.array(hist.axes.edges.T[1].flat)
-ebins = np.array(hist.axes.edges.T[2].flat)
+# ebins = np.array(hist.axes.edges.T[2].flat)
-p = Interpolator(eta, etabinsx[0:-1], etabinsy[0:-1], ebins[0:-1])
+# p = Interpolator(eta, etabinsx[0:-1], etabinsy[0:-1], ebins[0:-1])
 
 
 
 
-#Generate the hit
+# #Generate the hit
 
 
 
 
-tmp = p.interpolate(v)
+# tmp = p.interpolate(v)
-print(f'tmp:{tmp}')
+# print(f'tmp:{tmp}')
-pos = np.array((tmp['x'], tmp['y']))*25
+# pos = np.array((tmp['x'], tmp['y']))*25
 
 
-print(pixels)
+# print(pixels)
-fig, ax = plt.subplots(figsize = (7,7))
+# fig, ax = plt.subplots(figsize = (7,7))
-ax.pcolormesh(xaxis, xaxis, t)
+# ax.pcolormesh(xaxis, xaxis, t)
-ax.plot(*pos, 'o')
+# ax.plot(*pos, 'o')
-ax.set_xticks([0,25,50,75])
+# ax.set_xticks([0,25,50,75])
-ax.set_yticks([0,25,50,75])
+# ax.set_yticks([0,25,50,75])
-ax.set_xlim(0,75)
+# ax.set_xlim(0,75)
-ax.set_ylim(0,75)
+# ax.set_ylim(0,75)
-ax.grid()
+# ax.grid()
-print(f'{hit=}')
+# print(f'{hit=}')
-print(f'{pos=}')
+# print(f'{pos=}')

python/src/bind_Cluster.hpp

@@ -0,0 +1,64 @@
+#include "aare/Cluster.hpp"
+
+#include <cstdint>
+#include <filesystem>
+#include <fmt/format.h>
+#include <pybind11/numpy.h>
+#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>
+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;
+        }));
+
+    /*
+    //TODO! Review if to keep or not
+    .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!!!
+
+        });
+    */
+}
+
+#pragma GCC diagnostic pop

python/src/bind_ClusterCollector.hpp

@@ -0,0 +1,44 @@
+#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 T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType = uint16_t>
+void define_ClusterCollector(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<ClusterType> &self) {
+                auto v = new std::vector<ClusterVector<ClusterType>>(
+                    self.steal_clusters());
+                return v; // TODO change!!!
+            },
+            py::return_value_policy::take_ownership);
+}
+
+#pragma GCC diagnostic pop

python/src/bind_ClusterFile.hpp

@@ -21,8 +21,7 @@ using namespace ::aare;
 
 template <typename Type, uint8_t CoordSizeX, uint8_t CoordSizeY,
           typename CoordType = uint16_t>
-void define_cluster_file_io_bindings(py::module &m,
+void define_ClusterFile(py::module &m, const std::string &typestr) {
-                                     const std::string &typestr) {
 
     using ClusterType = Cluster<Type, CoordSizeX, CoordSizeY, CoordType>;
 
@@ -39,19 +38,20 @@ void define_cluster_file_io_bindings(py::module &m,
                     self.read_clusters(n_clusters));
                 return v;
             },
-            py::return_value_policy::take_ownership)
+            py::return_value_policy::take_ownership, py::arg("n_clusters"))
         .def("read_frame",
              [](ClusterFile<ClusterType> &self) {
                  auto v = new ClusterVector<ClusterType>(self.read_frame());
                  return v;
              })
-        .def("set_roi", &ClusterFile<ClusterType>::set_roi)
+        .def("set_roi", &ClusterFile<ClusterType>::set_roi,
+             py::arg("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);
-            })
+            }, py::arg("noise_map"))
 
         .def("set_gain_map",
              [](ClusterFile<ClusterType> &self, py::array_t<double> gain_map) {
@@ -59,9 +59,6 @@ void define_cluster_file_io_bindings(py::module &m,
                  self.set_gain_map(view);
              })
 
-        // void set_gain_map(const GainMap &gain_map); //TODO do i need a
-        // gainmap constructor?
-
         .def("close", &ClusterFile<ClusterType>::close)
         .def("write_frame", &ClusterFile<ClusterType>::write_frame)
         .def("__enter__", [](ClusterFile<ClusterType> &self) { return &self; })

python/src/bind_ClusterFileSink.hpp

@@ -0,0 +1,37 @@
+#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 T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType = uint16_t>
+void define_ClusterFileSink(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<ClusterType>::stop);
+}
+
+#pragma GCC diagnostic pop

python/src/bind_ClusterFinder.hpp

@@ -0,0 +1,77 @@
+#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 T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType = uint16_t>
+void define_ClusterFinder(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<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<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<ClusterType, uint16_t, pd_type> &self,
+               bool realloc_same_capacity) {
+                ClusterVector<ClusterType> clusters =
+                    self.steal_clusters(realloc_same_capacity);
+                return clusters;
+            },
+            py::arg("realloc_same_capacity") = false)
+        .def(
+            "find_clusters",
+            [](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);
+}
+
+#pragma GCC diagnostic pop

python/src/bind_ClusterFinderMT.hpp

@@ -0,0 +1,81 @@
+#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 T, uint8_t ClusterSizeX, uint8_t ClusterSizeY,
+          typename CoordType = uint16_t>
+void define_ClusterFinderMT(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<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<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_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);
+}
+
+#pragma GCC diagnostic pop

python/src/bind_ClusterVector.hpp

@@ -21,16 +21,14 @@ 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 =
+    using ClusterType = Cluster<Type, ClusterSizeX, ClusterSizeY, CoordType>;
-        Cluster<Type, ClusterSizeX, ClusterSizeY, CoordType, void>;
     auto class_name = fmt::format("ClusterVector_{}", typestr);
 
     py::class_<ClusterVector<
-        Cluster<Type, ClusterSizeX, ClusterSizeY, CoordType, void>, void>>(
+        Cluster<Type, ClusterSizeX, ClusterSizeY, CoordType>, void>>(
         m, class_name.c_str(),
         py::buffer_protocol())
 
@@ -41,10 +39,16 @@ void define_ClusterVector(py::module &m, const std::string &typestr) {
                  self.push_back(cluster);
              })
 
-        .def("sum", [](ClusterVector<ClusterType> &self) {
+        .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",
@@ -75,7 +79,6 @@ void define_ClusterVector(py::module &m, const std::string &typestr) {
     // 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
@@ -88,16 +91,27 @@ void define_ClusterVector(py::module &m, const std::string &typestr) {
                   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) {
+              for (const auto &cluster : cv) {
                   auto x = cluster.x;
                   auto y = cluster.y;
-                    if(x<image_size[1] && y<image_size[0])
+                  if (x < image_size[1] && y < image_size[0])
                       r(cluster.y, cluster.x) += 1;
               }
 
               return hitmap;
           });
 }
+
+void define_reduction(py::module &m) {
+    m.def("reduce_3x3_to_2x2", [](const ClusterVector<Cluster<int, 3, 3, uint16_t>> &cv) {
+        return new ClusterVector<Cluster<int, 2, 2, uint16_t>>(reduce_3x3_to_2x2(cv));
+        // return new ClusterVector<Cluster<int, 3, 3>>();
+    })
+    .def("reduce_5x5_to_3x3", [](const ClusterVector<Cluster<int, 5, 5, uint16_t>> &cv) {
+        return new ClusterVector<Cluster<int, 3, 3, uint16_t>>(reduce_5x5_to_3x3(cv));
+    });
+}
+
+#pragma GCC diagnostic pop

python/src/cluster.hpp

@@ -1,211 +0,0 @@
-#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>
-void define_cluster(py::module &m, const std::string &typestr) {
-    auto class_name = fmt::format("Cluster{}", typestr);
-
-    py::class_<Cluster<Type, ClusterSizeX, ClusterSizeY, CoordType, void>>(
-        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();
-            Type *ptr = static_cast<Type *>(buf_info.ptr);
-            Cluster<Type, ClusterSizeX, ClusterSizeY, CoordType, void> cluster;
-            cluster.x = x;
-            cluster.y = y;
-            std::copy(ptr, ptr + ClusterSizeX * ClusterSizeY,
-                      cluster.data); // Copy array contents
-            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!!!
-
-        });
-    */
-}
-
-
-
-
-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<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<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<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);
-}
-
-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<ClusterType> &self) {
-                auto v = new std::vector<ClusterVector<ClusterType>>(
-                    self.steal_clusters());
-                return v; // TODO change!!!
-            },
-            py::return_value_policy::take_ownership);
-}
-
-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<ClusterType>::stop);
-}
-
-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<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<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<ClusterType, uint16_t, pd_type> &self,
-               bool realloc_same_capacity) {
-                ClusterVector<ClusterType> clusters =
-                    self.steal_clusters(realloc_same_capacity);
-                return clusters;
-            },
-            py::arg("realloc_same_capacity") = false)
-        .def(
-            "find_clusters",
-            [](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);
-
-}
-#pragma GCC diagnostic pop

python/src/ctb_raw_file.hpp

@@ -6,10 +6,12 @@
 #include "aare/RawMasterFile.hpp"
 #include "aare/RawSubFile.hpp"
 
-#include "aare/defs.hpp"
 #include "aare/decode.hpp"
+#include "aare/defs.hpp"
 // #include "aare/fClusterFileV2.hpp"
 
+#include "np_helper.hpp"
+
 #include <cstdint>
 #include <filesystem>
 #include <pybind11/iostream.h>
@@ -24,46 +26,74 @@ using namespace ::aare;
 
 void define_ctb_raw_file_io_bindings(py::module &m) {
 
-m.def("adc_sar_05_decode64to16", [](py::array_t<uint8_t> input) {
+    m.def("adc_sar_05_decode64to16", [](py::array_t<uint8_t> input) {
-
+        if (input.ndim() != 2) {
-    
+            throw std::runtime_error(
-    if(input.ndim() != 2){
+                "Only 2D arrays are supported at this moment");
-        throw std::runtime_error("Only 2D arrays are supported at this moment"); 
         }
 
-    //Create a 2D output array with the same shape as the input
+        // Create a 2D output array with the same shape as the input
-    std::vector<ssize_t> shape{input.shape(0), input.shape(1)/static_cast<int64_t>(bits_per_byte)};
+        std::vector<ssize_t> shape{input.shape(0),
+                                   input.shape(1) /
+                                       static_cast<ssize_t>(bits_per_byte)};
         py::array_t<uint16_t> output(shape);
 
-    //Create a view of the input and output arrays
+        // Create a view of the input and output arrays
-    NDView<uint64_t, 2> input_view(reinterpret_cast<uint64_t*>(input.mutable_data()), {output.shape(0), output.shape(1)});
+        NDView<uint64_t, 2> input_view(
-    NDView<uint16_t, 2> output_view(output.mutable_data(), {output.shape(0), output.shape(1)});
+            reinterpret_cast<uint64_t *>(input.mutable_data()),
+            {output.shape(0), output.shape(1)});
+        NDView<uint16_t, 2> output_view(output.mutable_data(),
+                                        {output.shape(0), output.shape(1)});
 
         adc_sar_05_decode64to16(input_view, output_view);
 
         return output;
-});
+    });
 
-
+    m.def("adc_sar_04_decode64to16", [](py::array_t<uint8_t> input) {
-m.def("adc_sar_04_decode64to16", [](py::array_t<uint8_t> input) {
+        if (input.ndim() != 2) {
-
+            throw std::runtime_error(
-    
+                "Only 2D arrays are supported at this moment");
-    if(input.ndim() != 2){
-        throw std::runtime_error("Only 2D arrays are supported at this moment"); 
         }
 
-    //Create a 2D output array with the same shape as the input
+        // Create a 2D output array with the same shape as the input
-    std::vector<ssize_t> shape{input.shape(0), input.shape(1)/static_cast<int64_t>(bits_per_byte)};
+        std::vector<ssize_t> shape{input.shape(0),
+                                   input.shape(1) /
+                                       static_cast<ssize_t>(bits_per_byte)};
         py::array_t<uint16_t> output(shape);
 
-    //Create a view of the input and output arrays
+        // Create a view of the input and output arrays
-    NDView<uint64_t, 2> input_view(reinterpret_cast<uint64_t*>(input.mutable_data()), {output.shape(0), output.shape(1)});
+        NDView<uint64_t, 2> input_view(
-    NDView<uint16_t, 2> output_view(output.mutable_data(), {output.shape(0), output.shape(1)});
+            reinterpret_cast<uint64_t *>(input.mutable_data()),
+            {output.shape(0), output.shape(1)});
+        NDView<uint16_t, 2> output_view(output.mutable_data(),
+                                        {output.shape(0), output.shape(1)});
 
         adc_sar_04_decode64to16(input_view, output_view);
 
         return output;
-});
+    });
+
+    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) {
+              // 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);
+
+              // 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()});
+
+              apply_custom_weights(input_view, output_view, weights_view);
+              return output;
+          });
 
     py::class_<CtbRawFile>(m, "CtbRawFile")
         .def(py::init<const std::filesystem::path &>())
@@ -95,5 +125,4 @@ m.def("adc_sar_04_decode64to16", [](py::array_t<uint8_t> input) {
                                &CtbRawFile::image_size_in_bytes)
 
         .def_property_readonly("frames_in_file", &CtbRawFile::frames_in_file);
-
 }

python/src/file.hpp

@@ -20,14 +20,13 @@
 namespace py = pybind11;
 using namespace ::aare;
 
-//Disable warnings for unused parameters, as we ignore some
+// Disable warnings for unused parameters, as we ignore some
-//in the __exit__ method
+// in the __exit__ method
 #pragma GCC diagnostic push
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 
 void define_file_io_bindings(py::module &m) {
 
-
     py::enum_<DetectorType>(m, "DetectorType")
         .value("Jungfrau", DetectorType::Jungfrau)
         .value("Eiger", DetectorType::Eiger)
@@ -38,13 +37,10 @@ void define_file_io_bindings(py::module &m) {
         .value("ChipTestBoard", DetectorType::ChipTestBoard)
         .value("Unknown", DetectorType::Unknown);
 
-
     PYBIND11_NUMPY_DTYPE(DetectorHeader, frameNumber, expLength, packetNumber,
                          bunchId, timestamp, modId, row, column, reserved,
                          debug, roundRNumber, detType, version, packetMask);
 
-   
-
     py::class_<File>(m, "File")
         .def(py::init([](const std::filesystem::path &fname) {
             return File(fname, "r", {});
@@ -109,10 +105,12 @@ void define_file_io_bindings(py::module &m) {
                      reinterpret_cast<std::byte *>(image.mutable_data()));
                  return image;
              })
-        .def("read_n", [](File &self, size_t n_frames) {
+        .def("read_n",
-            //adjust for actual frames left in the file
+             [](File &self, size_t n_frames) {
-            n_frames = std::min(n_frames, self.total_frames()-self.tell());
+                 // adjust for actual frames left in the file
-            if(n_frames == 0){
+                 n_frames =
+                     std::min(n_frames, self.total_frames() - self.tell());
+                 if (n_frames == 0) {
                      throw std::runtime_error("No frames left in file");
                  }
                  std::vector<size_t> shape{n_frames, self.rows(), self.cols()};
@@ -126,22 +124,21 @@ void define_file_io_bindings(py::module &m) {
                  } else if (item_size == 4) {
                      image = py::array_t<uint32_t>(shape);
                  }
-            self.read_into(reinterpret_cast<std::byte *>(image.mutable_data()),
+                 self.read_into(
+                     reinterpret_cast<std::byte *>(image.mutable_data()),
                      n_frames);
                  return image;
              })
         .def("__enter__", [](File &self) { return &self; })
         .def("__exit__",
-             [](File &self,
+             [](File &self, const std::optional<pybind11::type> &exc_type,
-                const std::optional<pybind11::type> &exc_type,
                 const std::optional<pybind11::object> &exc_value,
                 const std::optional<pybind11::object> &traceback) {
                  //  self.close();
              })
         .def("__iter__", [](File &self) { return &self; })
         .def("__next__", [](File &self) {
-
+            try {
-            try{
                 const uint8_t item_size = self.bytes_per_pixel();
                 py::array image;
                 std::vector<ssize_t> shape;
@@ -158,12 +155,11 @@ void define_file_io_bindings(py::module &m) {
                 self.read_into(
                     reinterpret_cast<std::byte *>(image.mutable_data()));
                 return image;
-            }catch(std::runtime_error &e){
+            } catch (std::runtime_error &e) {
                 throw py::stop_iteration();
             }
         });
 
-
     py::class_<FileConfig>(m, "FileConfig")
         .def(py::init<>())
         .def_readwrite("rows", &FileConfig::rows)
@@ -180,8 +176,6 @@ void define_file_io_bindings(py::module &m) {
             return "<FileConfig: " + a.to_string() + ">";
         });
 
-
-
     py::class_<ScanParameters>(m, "ScanParameters")
         .def(py::init<const std::string &>())
         .def(py::init<const ScanParameters &>())
@@ -192,57 +186,27 @@ void define_file_io_bindings(py::module &m) {
         .def_property_readonly("stop", &ScanParameters::stop)
         .def_property_readonly("step", &ScanParameters::step);
 
-
     py::class_<ROI>(m, "ROI")
         .def(py::init<>())
-        .def(py::init<int64_t, int64_t, int64_t, int64_t>(), py::arg("xmin"),
+        .def(py::init<ssize_t, ssize_t, ssize_t, ssize_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)
         .def_readwrite("ymax", &ROI::ymax)
-        .def("__str__", [](const ROI& self){
+        .def("__str__",
-            return fmt::format("ROI: xmin: {} xmax: {} ymin: {} ymax: {}", self.xmin, self.xmax, self.ymin, self.ymax);
+             [](const ROI &self) {
+                 return fmt::format("ROI: xmin: {} xmax: {} ymin: {} ymax: {}",
+                                    self.xmin, self.xmax, self.ymin, self.ymax);
              })
-        .def("__repr__", [](const ROI& self){
+        .def("__repr__",
-            return fmt::format("<ROI: xmin: {} xmax: {} ymin: {} ymax: {}>", self.xmin, self.xmax, self.ymin, self.ymax);
+             [](const ROI &self) {
+                 return fmt::format(
+                     "<ROI: xmin: {} xmax: {} ymin: {} ymax: {}>", self.xmin,
+                     self.xmax, self.ymin, self.ymax);
              })
         .def("__iter__", [](const ROI &self) {
-            return py::make_iterator(&self.xmin, &self.ymax+1); //NOLINT
+            return py::make_iterator(&self.xmin, &self.ymax + 1); // NOLINT
-        });
-
-    
-
-
-    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

python/src/fit.hpp

@@ -9,7 +9,6 @@
 namespace py = pybind11;
 using namespace pybind11::literals;
 
-
 void define_fit_bindings(py::module &m) {
 
     // TODO! Evaluate without converting to double
@@ -55,6 +54,49 @@ void define_fit_bindings(py::module &m) {
         )",
         py::arg("x"), py::arg("par"));
 
+    m.def(
+        "scurve",
+        [](py::array_t<double, py::array::c_style | py::array::forcecast> x,
+           py::array_t<double, py::array::c_style | py::array::forcecast> par) {
+            auto x_view = make_view_1d(x);
+            auto par_view = make_view_1d(par);
+            auto y =
+                new NDArray<double, 1>{aare::func::scurve(x_view, par_view)};
+            return return_image_data(y);
+        },
+        R"(
+        Evaluate a 1D scurve function for all points in x using parameters par. 
+        
+        Parameters
+        ----------
+        x : array_like
+            The points at which to evaluate the scurve function.
+        par : array_like
+            The parameters of the scurve function. The first element is the background slope, the second element is the background intercept, the third element is the mean, the fourth element is the standard deviation, the fifth element is inflexion point count number, and the sixth element is C.     
+        )",
+        py::arg("x"), py::arg("par"));
+
+    m.def(
+        "scurve2",
+        [](py::array_t<double, py::array::c_style | py::array::forcecast> x,
+           py::array_t<double, py::array::c_style | py::array::forcecast> par) {
+            auto x_view = make_view_1d(x);
+            auto par_view = make_view_1d(par);
+            auto y =
+                new NDArray<double, 1>{aare::func::scurve2(x_view, par_view)};
+            return return_image_data(y);
+        },
+        R"(
+        Evaluate a 1D scurve2 function for all points in x using parameters par. 
+        
+        Parameters
+        ----------
+        x : array_like
+            The points at which to evaluate the scurve function.
+        par : array_like
+            The parameters of the scurve2 function. The first element is the background slope, the second element is the background intercept, the third element is the mean, the fourth element is the standard deviation, the fifth element is inflexion point count number, and the sixth element is C.     
+        )",
+        py::arg("x"), py::arg("par"));
 
     m.def(
         "fit_gaus",
@@ -98,7 +140,6 @@ n_threads : int, optional
            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) {
-
             if (y.ndim() == 3) {
                 // Allocate memory for the output
                 // Need to have pointers to allow python to manage
@@ -132,7 +173,6 @@ 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(), chi2);
@@ -211,7 +251,6 @@ n_threads : int, optional
                                 "chi2"_a = return_image_data(chi2),
                                 "Ndf"_a = y.shape(2) - 2);
 
-
             } else if (y.ndim() == 1) {
                 auto par = new NDArray<double, 1>({2});
                 auto par_err = new NDArray<double, 1>({2});
@@ -235,6 +274,177 @@ n_threads : int, optional
         R"(
 Fit a 1D polynomial to data with error estimates.
 
+Parameters
+----------
+x : array_like
+    The x values.
+y : array_like
+    The y values.
+y_err : array_like
+    The error in the y values.
+n_threads : int, optional
+    The number of threads to use. Default is 4.
+)",
+        py::arg("x"), py::arg("y"), py::arg("y_err"), py::arg("n_threads") = 4);
+
+    //=========
+    m.def(
+        "fit_scurve",
+        [](py::array_t<double, py::array::c_style | py::array::forcecast> x,
+           py::array_t<double, py::array::c_style | py::array::forcecast> y,
+           int n_threads) {
+            if (y.ndim() == 3) {
+                auto par = new NDArray<double, 3>{};
+
+                auto x_view = make_view_1d(x);
+                auto y_view = make_view_3d(y);
+                *par = aare::fit_scurve(x_view, y_view, n_threads);
+                return return_image_data(par);
+            } else if (y.ndim() == 1) {
+                auto par = new NDArray<double, 1>{};
+                auto x_view = make_view_1d(x);
+                auto y_view = make_view_1d(y);
+                *par = aare::fit_scurve(x_view, y_view);
+                return return_image_data(par);
+            } else {
+                throw std::runtime_error("Data must be 1D or 3D");
+            }
+        },
+        py::arg("x"), py::arg("y"), py::arg("n_threads") = 4);
+
+    m.def(
+        "fit_scurve",
+        [](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) {
+            if (y.ndim() == 3) {
+                auto par = new NDArray<double, 3>({y.shape(0), y.shape(1), 6});
+
+                auto par_err =
+                    new NDArray<double, 3>({y.shape(0), y.shape(1), 6});
+
+                auto y_view = make_view_3d(y);
+                auto y_view_err = make_view_3d(y_err);
+                auto x_view = make_view_1d(x);
+
+                auto chi2 = new NDArray<double, 2>({y.shape(0), y.shape(1)});
+
+                aare::fit_scurve(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});
+                auto par_err = new NDArray<double, 1>({2});
+
+                auto y_view = make_view_1d(y);
+                auto y_view_err = make_view_1d(y_err);
+                auto x_view = make_view_1d(x);
+
+                double chi2 = 0;
+
+                aare::fit_scurve(x_view, y_view, y_view_err, par->view(),
+                                 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"(
+Fit a 1D polynomial to data with error estimates.
+
+Parameters
+----------
+x : array_like
+    The x values.
+y : array_like
+    The y values.
+y_err : array_like
+    The error in the y values.
+n_threads : int, optional
+    The number of threads to use. Default is 4.
+)",
+        py::arg("x"), py::arg("y"), py::arg("y_err"), py::arg("n_threads") = 4);
+
+    m.def(
+        "fit_scurve2",
+        [](py::array_t<double, py::array::c_style | py::array::forcecast> x,
+           py::array_t<double, py::array::c_style | py::array::forcecast> y,
+           int n_threads) {
+            if (y.ndim() == 3) {
+                auto par = new NDArray<double, 3>{};
+
+                auto x_view = make_view_1d(x);
+                auto y_view = make_view_3d(y);
+                *par = aare::fit_scurve2(x_view, y_view, n_threads);
+                return return_image_data(par);
+            } else if (y.ndim() == 1) {
+                auto par = new NDArray<double, 1>{};
+                auto x_view = make_view_1d(x);
+                auto y_view = make_view_1d(y);
+                *par = aare::fit_scurve2(x_view, y_view);
+                return return_image_data(par);
+            } else {
+                throw std::runtime_error("Data must be 1D or 3D");
+            }
+        },
+        py::arg("x"), py::arg("y"), py::arg("n_threads") = 4);
+
+    m.def(
+        "fit_scurve2",
+        [](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) {
+            if (y.ndim() == 3) {
+                auto par = new NDArray<double, 3>({y.shape(0), y.shape(1), 6});
+
+                auto par_err =
+                    new NDArray<double, 3>({y.shape(0), y.shape(1), 6});
+
+                auto y_view = make_view_3d(y);
+                auto y_view_err = make_view_3d(y_err);
+                auto x_view = make_view_1d(x);
+
+                auto chi2 = new NDArray<double, 2>({y.shape(0), y.shape(1)});
+
+                aare::fit_scurve2(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>({6});
+                auto par_err = new NDArray<double, 1>({6});
+
+                auto y_view = make_view_1d(y);
+                auto y_view_err = make_view_1d(y_err);
+                auto x_view = make_view_1d(x);
+
+                double chi2 = 0;
+
+                aare::fit_scurve2(x_view, y_view, y_view_err, par->view(),
+                                  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"(
+Fit a 1D polynomial to data with error estimates.
+
 Parameters
 ----------
 x : array_like

python/src/jungfrau_data_file.hpp

@@ -21,10 +21,7 @@ using namespace ::aare;
 
 auto read_dat_frame(JungfrauDataFile &self) {
     py::array_t<JungfrauDataHeader> header(1);
-    py::array_t<uint16_t> image({
+    py::array_t<uint16_t> image({self.rows(), self.cols()});
-        self.rows(), 
-        self.cols() 
-    });
 
     self.read_into(reinterpret_cast<std::byte *>(image.mutable_data()),
                    header.mutable_data());
@@ -40,9 +37,7 @@ auto read_n_dat_frames(JungfrauDataFile &self, size_t n_frames) {
     }
 
     py::array_t<JungfrauDataHeader> header(n_frames);
-    py::array_t<uint16_t> image({
+    py::array_t<uint16_t> image({n_frames, self.rows(), self.cols()});
-        n_frames, self.rows(), 
-        self.cols()});
 
     self.read_into(reinterpret_cast<std::byte *>(image.mutable_data()),
                    n_frames, header.mutable_data());

python/src/module.cpp

@@ -1,32 +1,57 @@
 // Files with bindings to the different classes
 
-//New style file naming
+// New style file naming
+#include "bind_Cluster.hpp"
+#include "bind_ClusterCollector.hpp"
+#include "bind_ClusterFile.hpp"
+#include "bind_ClusterFileSink.hpp"
+#include "bind_ClusterFinder.hpp"
+#include "bind_ClusterFinderMT.hpp"
 #include "bind_ClusterVector.hpp"
 
-//TODO! migrate the other names
+// 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 "jungfrau_data_file.hpp"
 #include "pedestal.hpp"
 #include "pixel_map.hpp"
 #include "raw_file.hpp"
 #include "raw_master_file.hpp"
+#include "raw_sub_file.hpp"
 #include "var_cluster.hpp"
 
-#include "jungfrau_data_file.hpp"
-
 // Pybind stuff
 #include <pybind11/pybind11.h>
 #include <pybind11/stl.h>
 
 namespace py = pybind11;
 
+/* MACRO that defines Cluster bindings for a specific size and type
+
+T - Storage type of the cluster data (int, float, double)
+N - Number of rows in the cluster
+M - Number of columns in the cluster
+U - Type of the pixel data (e.g., uint16_t)
+TYPE_CODE - A character representing the type code (e.g., 'i' for int, 'd' for
+double, 'f' for float)
+
+*/
+#define DEFINE_CLUSTER_BINDINGS(T, N, M, U, TYPE_CODE)                         \
+    define_ClusterFile<T, N, M, U>(m, "Cluster" #N "x" #M #TYPE_CODE);         \
+    define_ClusterVector<T, N, M, U>(m, "Cluster" #N "x" #M #TYPE_CODE);       \
+    define_ClusterFinder<T, N, M, U>(m, "Cluster" #N "x" #M #TYPE_CODE);       \
+    define_ClusterFinderMT<T, N, M, U>(m, "Cluster" #N "x" #M #TYPE_CODE);     \
+    define_ClusterFileSink<T, N, M, U>(m, "Cluster" #N "x" #M #TYPE_CODE);     \
+    define_ClusterCollector<T, N, M, U>(m, "Cluster" #N "x" #M #TYPE_CODE);    \
+    define_Cluster<T, N, M, U>(m, #N "x" #M #TYPE_CODE);                       \
+    register_calculate_eta<T, N, M, U>(m);
+
 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);
@@ -37,59 +62,27 @@ PYBIND11_MODULE(_aare, 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_BINDINGS(int, 3, 3, uint16_t, i);
-    define_cluster_file_io_bindings<double, 3, 3, uint16_t>(m, "Cluster3x3d");
+    DEFINE_CLUSTER_BINDINGS(double, 3, 3, uint16_t, d);
-    define_cluster_file_io_bindings<float, 3, 3, uint16_t>(m, "Cluster3x3f");
+    DEFINE_CLUSTER_BINDINGS(float, 3, 3, uint16_t, f);
-    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_CLUSTER_BINDINGS(int, 2, 2, uint16_t, i);
-    define_ClusterVector<double, 3, 3, uint16_t>(m, "Cluster3x3d");
+    DEFINE_CLUSTER_BINDINGS(double, 2, 2, uint16_t, d);
-    define_ClusterVector<float, 3, 3, uint16_t>(m, "Cluster3x3f");
+    DEFINE_CLUSTER_BINDINGS(float, 2, 2, uint16_t, f);
-    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_BINDINGS(int, 5, 5, uint16_t, i);
-    define_cluster_finder_bindings<double, 3, 3, uint16_t>(m, "Cluster3x3d");
+    DEFINE_CLUSTER_BINDINGS(double, 5, 5, uint16_t, d);
-    define_cluster_finder_bindings<float, 3, 3, uint16_t>(m, "Cluster3x3f");
+    DEFINE_CLUSTER_BINDINGS(float, 5, 5, uint16_t, f);
-    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_BINDINGS(int, 7, 7, uint16_t, i);
-    define_cluster_finder_mt_bindings<double, 3, 3, uint16_t>(m, "Cluster3x3d");
+    DEFINE_CLUSTER_BINDINGS(double, 7, 7, uint16_t, d);
-    define_cluster_finder_mt_bindings<float, 3, 3, uint16_t>(m, "Cluster3x3f");
+    DEFINE_CLUSTER_BINDINGS(float, 7, 7, uint16_t, f);
-    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_BINDINGS(int, 9, 9, uint16_t, i);
-    define_cluster_file_sink_bindings<double, 3, 3, uint16_t>(m, "Cluster3x3d");
+    DEFINE_CLUSTER_BINDINGS(double, 9, 9, uint16_t, d);
-    define_cluster_file_sink_bindings<float, 3, 3, uint16_t>(m, "Cluster3x3f");
+    DEFINE_CLUSTER_BINDINGS(float, 9, 9, uint16_t, f);
-    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_reduction(m);
-    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

@@ -13,7 +13,7 @@ namespace py = pybind11;
 using namespace aare;
 
 // Pass image data back to python as a numpy array
-template <typename T, int64_t Ndim>
+template <typename T, ssize_t Ndim>
 py::array return_image_data(aare::NDArray<T, Ndim> *image) {
 
     py::capsule free_when_done(image, [](void *f) {

python/src/pedestal.hpp

@@ -9,7 +9,8 @@
 
 namespace py = pybind11;
 
-template <typename SUM_TYPE> void define_pedestal_bindings(py::module &m, const std::string &name) {
+template <typename SUM_TYPE>
+void define_pedestal_bindings(py::module &m, const std::string &name) {
     py::class_<Pedestal<SUM_TYPE>>(m, name.c_str())
         .def(py::init<int, int, int>())
         .def(py::init<int, int>())
@@ -19,12 +20,14 @@ template <typename SUM_TYPE> void define_pedestal_bindings(py::module &m, const
                  *mea = self.mean();
                  return return_image_data(mea);
              })
-        .def("variance", [](Pedestal<SUM_TYPE> &self) {
+        .def("variance",
+             [](Pedestal<SUM_TYPE> &self) {
                  auto var = new NDArray<SUM_TYPE, 2>{};
                  *var = self.variance();
                  return return_image_data(var);
              })
-        .def("std", [](Pedestal<SUM_TYPE> &self) {
+        .def("std",
+             [](Pedestal<SUM_TYPE> &self) {
                  auto std = new NDArray<SUM_TYPE, 2>{};
                  *std = self.std();
                  return return_image_data(std);
@@ -39,14 +42,19 @@ template <typename SUM_TYPE> void define_pedestal_bindings(py::module &m, const
              [&](Pedestal<SUM_TYPE> &pedestal) {
                  return Pedestal<SUM_TYPE>(pedestal);
              })
-        //TODO! add push for other data types
+        // TODO! add push for other data types
-        .def("push", [](Pedestal<SUM_TYPE> &pedestal, py::array_t<uint16_t> &f) {
+        .def("push",
+             [](Pedestal<SUM_TYPE> &pedestal, py::array_t<uint16_t> &f) {
                  auto v = make_view_2d(f);
                  pedestal.push(v);
              })
-        .def("push_no_update", [](Pedestal<SUM_TYPE> &pedestal, py::array_t<uint16_t, py::array::c_style> &f) {
+        .def(
+            "push_no_update",
+            [](Pedestal<SUM_TYPE> &pedestal,
+               py::array_t<uint16_t, py::array::c_style> &f) {
                 auto v = make_view_2d(f);
                 pedestal.push_no_update(v);
-        }, py::arg().noconvert())
+            },
+            py::arg().noconvert())
         .def("update_mean", &Pedestal<SUM_TYPE>::update_mean);
 }

python/src/pixel_map.hpp

@@ -1,41 +1,46 @@
 #include "aare/PixelMap.hpp"
 #include "np_helper.hpp"
 
-
 #include <cstdint>
 #include <pybind11/numpy.h>
 #include <pybind11/pybind11.h>
 #include <pybind11/stl.h>
 
-
 namespace py = pybind11;
-using namespace::aare;
+using namespace ::aare;
-
 
 void define_pixel_map_bindings(py::module &m) {
-    m.def("GenerateMoench03PixelMap", []() {
+    m.def("GenerateMoench03PixelMap",
-        auto ptr = new NDArray<ssize_t,2>(GenerateMoench03PixelMap());
+          []() {
+              auto ptr = new NDArray<ssize_t, 2>(GenerateMoench03PixelMap());
               return return_image_data(ptr);
           })
-    .def("GenerateMoench05PixelMap", []() {
+        .def("GenerateMoench05PixelMap",
-        auto ptr = new NDArray<ssize_t,2>(GenerateMoench05PixelMap());
+             []() {
+                 auto ptr = new NDArray<ssize_t, 2>(GenerateMoench05PixelMap());
                  return return_image_data(ptr);
              })
-        .def("GenerateMoench05PixelMap1g", []() {
+        .def("GenerateMoench05PixelMap1g",
-        auto ptr = new NDArray<ssize_t,2>(GenerateMoench05PixelMap1g());
+             []() {
+                 auto ptr =
+                     new NDArray<ssize_t, 2>(GenerateMoench05PixelMap1g());
                  return return_image_data(ptr);
              })
-    .def("GenerateMoench05PixelMapOld", []() {
+        .def("GenerateMoench05PixelMapOld",
-        auto ptr = new NDArray<ssize_t,2>(GenerateMoench05PixelMapOld());
+             []() {
+                 auto ptr =
+                     new NDArray<ssize_t, 2>(GenerateMoench05PixelMapOld());
                  return return_image_data(ptr);
              })
-    .def("GenerateMH02SingleCounterPixelMap", []() {
+        .def("GenerateMH02SingleCounterPixelMap",
-        auto ptr = new NDArray<ssize_t,2>(GenerateMH02SingleCounterPixelMap());
+             []() {
+                 auto ptr = new NDArray<ssize_t, 2>(
+                     GenerateMH02SingleCounterPixelMap());
                  return return_image_data(ptr);
              })
         .def("GenerateMH02FourCounterPixelMap", []() {
-        auto ptr = new NDArray<ssize_t,3>(GenerateMH02FourCounterPixelMap());
+            auto ptr =
+                new NDArray<ssize_t, 3>(GenerateMH02FourCounterPixelMap());
             return return_image_data(ptr);
         });
-
 }

python/src/raw_file.hpp

@@ -32,7 +32,7 @@ void define_raw_file_io_bindings(py::module &m) {
                  shape.push_back(self.cols());
 
                  // return headers from all subfiles
-                 py::array_t<DetectorHeader> header(self.n_mod());
+                 py::array_t<DetectorHeader> header(self.n_modules());
 
                  const uint8_t item_size = self.bytes_per_pixel();
                  if (item_size == 1) {
@@ -61,10 +61,11 @@ void define_raw_file_io_bindings(py::module &m) {
 
                 // return headers from all subfiles
                 py::array_t<DetectorHeader> header;
-                if (self.n_mod() == 1) {
+                if (self.n_modules() == 1) {
                     header = py::array_t<DetectorHeader>(n_frames);
                 } else {
-                    header = py::array_t<DetectorHeader>({self.n_mod(), n_frames});
+                    header = py::array_t<DetectorHeader>(
+                        {self.n_modules_in_roi(), n_frames});
                 }
                 // py::array_t<DetectorHeader> header({self.n_mod(), n_frames});
 
@@ -100,7 +101,8 @@ void define_raw_file_io_bindings(py::module &m) {
         .def_property_readonly("cols", &RawFile::cols)
         .def_property_readonly("bitdepth", &RawFile::bitdepth)
         .def_property_readonly("geometry", &RawFile::geometry)
-        .def_property_readonly("n_mod", &RawFile::n_mod)
         .def_property_readonly("detector_type", &RawFile::detector_type)
-        .def_property_readonly("master", &RawFile::master);
+        .def_property_readonly("master", &RawFile::master)
+        .def_property_readonly("n_modules", &RawFile::n_modules)
+        .def_property_readonly("n_modules_in_roi", &RawFile::n_modules_in_roi);
 }

python/src/raw_master_file.hpp

@@ -57,7 +57,10 @@ void define_raw_master_file_bindings(py::module &m) {
         .def_property_readonly("total_frames_expected",
                                &RawMasterFile::total_frames_expected)
         .def_property_readonly("geometry", &RawMasterFile::geometry)
-        .def_property_readonly("analog_samples", &RawMasterFile::analog_samples, R"(
+        .def_property_readonly("udp_interfaces_per_module",
+                               &RawMasterFile::udp_interfaces_per_module)
+        .def_property_readonly("analog_samples", &RawMasterFile::analog_samples,
+                               R"(
             Number of analog samples
 
             Returns

python/src/raw_sub_file.hpp

@@ -0,0 +1,105 @@
+#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

@@ -12,10 +12,8 @@
 // #include <pybind11/stl/filesystem.h>
 // #include <string>
 
-
 namespace py = pybind11;
-using namespace::aare;
+using namespace ::aare;
-
 
 void define_var_cluster_finder_bindings(py::module &m) {
     PYBIND11_NUMPY_DTYPE(VarClusterFinder<double>::Hit, size, row, col,
@@ -65,9 +63,7 @@ void define_var_cluster_finder_bindings(py::module &m) {
                  return return_vector(ptr);
              })
         .def("clear_hits",
-             [](VarClusterFinder<double> &self) {
+             [](VarClusterFinder<double> &self) { self.clear_hits(); })
-                 self.clear_hits();
-             })
         .def("steal_hits",
              [](VarClusterFinder<double> &self) {
                  auto ptr = new std::vector<VarClusterFinder<double>::Hit>(
@@ -75,5 +71,4 @@ void define_var_cluster_finder_bindings(py::module &m) {
                  return return_vector(ptr);
              })
         .def("total_clusters", &VarClusterFinder<double>::total_clusters);
-
 }

python/tests/conftest.py

@@ -6,20 +6,20 @@ import pytest
 
 def pytest_addoption(parser):
     parser.addoption(
-        "--files", action="store_true", default=False, help="run slow tests"
+        "--with-data", action="store_true", default=False, help="Run tests that require additional data"
     )
 
 
 def pytest_configure(config):
-    config.addinivalue_line("markers", "files: mark test as needing image files to run")
+    config.addinivalue_line("markers", "withdata: mark test as needing image files to run")
 
 
 def pytest_collection_modifyitems(config, items):
-    if config.getoption("--files"):
+    if config.getoption("--with-data"):
         return
-    skip = pytest.mark.skip(reason="need --files option to run")
+    skip = pytest.mark.skip(reason="need --with-data option to run")
     for item in items:
-        if "files" in item.keywords:
+        if "withdata" in item.keywords:
             item.add_marker(skip)
 
 

python/tests/test_ClusterFile.py

@@ -9,7 +9,7 @@ import pickle
 from aare import ClusterFile
 from conftest import test_data_path
 
-@pytest.mark.files
+@pytest.mark.withdata
 def test_cluster_file(test_data_path): 
     """Test ClusterFile""" 
     f =  ClusterFile(test_data_path / "clust/single_frame_97_clustrers.clust") 
@@ -39,7 +39,7 @@ def test_cluster_file(test_data_path):
     for i in range(10):
         assert arr[i]['x'] == i+1
 
-@pytest.mark.files
+@pytest.mark.withdata
 def test_read_clusters_and_fill_histogram(test_data_path): 
     # Create the histogram
     n_bins = 100

python/tests/test_RawFile.py

@@ -0,0 +1,73 @@
+import pytest
+from aare import RawFile
+import numpy as np
+
+@pytest.mark.withdata
+def test_read_rawfile_with_roi(test_data_path):
+
+   with RawFile(test_data_path / "raw/SingleChipROI/Data_master_0.json") as f:
+    headers, frames = f.read()
+
+    assert headers.size == 10100
+    assert frames.shape == (10100, 256, 256)
+    
+@pytest.mark.withdata
+def test_read_rawfile_quad_eiger_and_compare_to_numpy(test_data_path): 
+    
+    d0 = test_data_path/'raw/eiger_quad_data/W13_vrpreampscan_m21C_300V_800eV_vthre2000_d0_f0_0.raw'
+    d1 = test_data_path/'raw/eiger_quad_data/W13_vrpreampscan_m21C_300V_800eV_vthre2000_d1_f0_0.raw'
+
+    image = np.zeros((512,512), dtype=np.uint32)
+
+    with open(d0) as f:
+        raw = np.fromfile(f, dtype=np.uint32, count = 256*512, offset = 20*256*512*4 + 112*21).reshape(256,512)
+
+        image[256:,:] = raw
+
+    with open(d1) as f:
+        raw = np.fromfile(f, dtype=np.uint32, count = 256*512, offset = 20*256*512*4 + 112*21).reshape(256,512)
+        
+        image[0:256,:] = raw[::-1,:]
+
+    with RawFile(test_data_path/'raw/eiger_quad_data/W13_vrpreampscan_m21C_300V_800eV_vthre2000_master_0.json') as f:
+        f.seek(20)
+        header, image1 = f.read_frame()
+    
+    assert (image == image1).all()
+
+
+@pytest.mark.withdata
+def test_read_rawfile_eiger_and_compare_to_numpy(test_data_path): 
+    d0 = test_data_path/'raw/eiger/Lab6_20500eV_2deg_20240629_d0_f0_7.raw'
+    d1 = test_data_path/'raw/eiger/Lab6_20500eV_2deg_20240629_d1_f0_7.raw'
+    d2 = test_data_path/'raw/eiger/Lab6_20500eV_2deg_20240629_d2_f0_7.raw'
+    d3 = test_data_path/'raw/eiger/Lab6_20500eV_2deg_20240629_d3_f0_7.raw'
+
+    image = np.zeros((512,1024), dtype=np.uint32)
+
+    #TODO why is there no header offset?
+    with open(d0) as f:
+        raw = np.fromfile(f, dtype=np.uint32, count = 256*512, offset=112).reshape(256,512)
+
+        image[0:256,0:512] = raw[::-1]
+
+    with open(d1) as f:
+        raw = np.fromfile(f, dtype=np.uint32, count = 256*512, offset=112).reshape(256,512)
+        
+        image[0:256,512:] = raw[::-1]
+
+    with open(d2) as f:
+        raw = np.fromfile(f, dtype=np.uint32, count = 256*512, offset=112).reshape(256,512)
+        
+        image[256:,0:512] = raw
+
+    with open(d3) as f:
+        raw = np.fromfile(f, dtype=np.uint32, count = 256*512, offset=112).reshape(256,512)
+        
+        image[256:,512:] = raw
+
+    
+    with RawFile(test_data_path/'raw/eiger/Lab6_20500eV_2deg_20240629_master_7.json') as f:
+        header, image1 = f.read_frame()
+
+    assert (image == image1).all()

python/tests/test_RawSubFile.py

@@ -0,0 +1,39 @@
+import pytest
+import numpy as np
+from aare import RawSubFile, DetectorType
+
+
+@pytest.mark.withdata
+def test_read_a_jungfrau_RawSubFile(test_data_path):
+
+    # Starting with f1 there is now 7 frames left in the series of files
+    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 == 7
+
+        headers, frames = f.read()
+
+    assert headers.size == 7
+    assert frames.shape == (7, 512, 1024)
+    
+
+    for i,h in zip(range(4,11,1), 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:] == frames)
+
+@pytest.mark.withdata
+def test_iterate_over_a_jungfrau_RawSubFile(test_data_path):
+
+    data = np.load(test_data_path / "raw/jungfrau/jungfrau_single_0.npy")
+
+    # Given the first subfile in a series we can read all frames from f0, f1, f2...fN
+    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 == 10
+        assert header["frameNumber"] == 10

python/tests/test_jungfrau_dat_files.py

@@ -2,7 +2,7 @@ import pytest
 import numpy as np
 from aare import JungfrauDataFile
 
-@pytest.mark.files
+@pytest.mark.withdata
 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
@@ -14,7 +14,7 @@ def test_jfungfrau_dat_read_number_of_frames(test_data_path):
         assert dat_file.total_frames == 113
 
 
-@pytest.mark.files
+@pytest.mark.withdata
 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
@@ -26,7 +26,7 @@ def test_jfungfrau_dat_read_number_of_file(test_data_path):
         assert dat_file.n_files == 7
 
 
-@pytest.mark.files
+@pytest.mark.withdata
 def test_read_module(test_data_path):
     """
     Read all frames from the series of .dat files. Compare to canned data in npz format. 
@@ -50,7 +50,7 @@ def test_read_module(test_data_path):
     assert np.all(ref_header == header)
     assert np.all(ref_data == data)
 
-@pytest.mark.files
+@pytest.mark.withdata
 def test_read_half_module(test_data_path):
 
     # Read all frames from the .dat file
@@ -71,7 +71,7 @@ def test_read_half_module(test_data_path):
     assert np.all(ref_data == data)
 
 
-@pytest.mark.files
+@pytest.mark.withdata
 def test_read_single_chip(test_data_path):
 
     # Read all frames from the .dat file

src/CalculateEta.test.cpp

@@ -21,20 +21,22 @@ using ClusterTypes =
 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, corner::cBottomLeft, 7}),
+                        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, corner::cTopRight, 20}),
+            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, 8, 8, 9, 2,
+                            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>{9. / 17, 5. / 13, 8, 28}),
+                        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>{7. / 11, 6. / 10, 1, 21}),
+            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, 4. / 6, 1, 11}));
+            Eta2<int>{3. / 5, 2. / 5, 1, 11}));
 }
 
 TEST_CASE("compute_largest_2x2_subcluster", "[eta_calculation]") {
@@ -61,14 +63,13 @@ TEST_CASE("calculate_eta2", "[eta_calculation]") {
     CHECK(eta.sum == expected_eta.sum);
 }
 
-
+// 3x3 cluster layout (rotated to match the cBottomLeft enum):
-//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 "
-TEST_CASE("Calculate eta2 for a 3x3 int32 cluster with the largest 2x2 sum in the bottom left",
+          "the bottom left",
           "[eta_calculation]") {
 
     // Create a 3x3 cluster
@@ -90,39 +91,37 @@ TEST_CASE("Calculate eta2 for a 3x3 int32 cluster with the largest 2x2 sum in th
     // 30, 23, 5
 
     auto eta = calculate_eta2(cl);
-    CHECK(eta.c == corner::cBottomLeft);
+    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);
+    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",
+TEST_CASE("Calculate eta2 for a 3x3 int32 cluster with the largest 2x2 sum in "
+          "the top left",
           "[eta_calculation]") {
 
-// Create a 3x3 cluster
+    // Create a 3x3 cluster
-Cluster<int32_t, 3, 3> cl;
+    Cluster<int32_t, 3, 3> cl;
-cl.x = 0;
+    cl.x = 0;
-cl.y = 0;
+    cl.y = 0;
-cl.data[0] = 8;
+    cl.data[0] = 8;
-cl.data[1] = 12;
+    cl.data[1] = 12;
-cl.data[2] = 5;
+    cl.data[2] = 5;
-cl.data[3] = 77;
+    cl.data[3] = 77;
-cl.data[4] = 80;
+    cl.data[4] = 80;
-cl.data[5] = 3;
+    cl.data[5] = 3;
-cl.data[6] = 82;
+    cl.data[6] = 82;
-cl.data[7] = 91;
+    cl.data[7] = 91;
-cl.data[8] = 3;
+    cl.data[8] = 3;
 
-// 82, 91, 3
+    // 82, 91, 3
-// 77, 80, 3
+    // 77, 80, 3
-// 8, 12, 5
+    // 8, 12, 5
-
-auto eta = calculate_eta2(cl);
-CHECK(eta.c == 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);
 
+    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

@@ -14,19 +14,6 @@
 
 using namespace aare;
 
-TEST_CASE("Correct Instantiation of Cluster and ClusterVector",
-          "[.cluster][.instantiation]") {
-
-    CHECK(is_valid_cluster<double, 3, 3>);
-    CHECK(is_valid_cluster<double, 3, 2>);
-    CHECK(not is_valid_cluster<int, 0, 0>);
-    CHECK(not is_valid_cluster<std::string, 2, 2>);
-    CHECK(not is_valid_cluster<int, 2, 2, double>);
-
-    CHECK(not is_cluster_v<int>);
-    CHECK(is_cluster_v<Cluster<int, 3, 3>>);
-}
-
 TEST_CASE("Test sum of Cluster", "[.cluster]") {
     Cluster<int, 2, 2> cluster{0, 0, {1, 2, 3, 4}};
 

src/ClusterFile.cpp

@@ -0,0 +1,395 @@
+#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);
+    }
+}
+
+void ClusterFile::set_roi(ROI roi) { m_roi = roi; }
+
+void ClusterFile::set_noise_map(const NDView<int32_t, 2> noise_map) {
+    m_noise_map = NDArray<int32_t, 2>(noise_map);
+}
+
+void ClusterFile::set_gain_map(const NDView<double, 2> gain_map) {
+    m_gain_map = NDArray<double, 2>(gain_map);
+
+    // Gain map is passed as ADU/keV to avoid dividing in when applying the gain
+    // map we invert it here
+    for (auto &item : m_gain_map->view()) {
+        item = 1.0 / item;
+    }
+}
+
+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");
+    }
+    // First write the frame number - 4 bytes
+    int32_t frame_number = clusters.frame_number();
+    if (fwrite(&frame_number, sizeof(frame_number), 1, fp) != 1) {
+        throw std::runtime_error(LOCATION + "Could not write frame number");
+    }
+
+    // Then write the number of clusters - 4 bytes
+    uint32_t n_clusters = clusters.size();
+    if (fwrite(&n_clusters, sizeof(n_clusters), 1, fp) != 1) {
+        throw std::runtime_error(LOCATION +
+                                 "Could not write number of clusters");
+    }
+
+    // Now write the clusters in the frame
+    if (fwrite(clusters.data(), clusters.item_size(), clusters.size(), fp) !=
+        clusters.size()) {
+        throw std::runtime_error(LOCATION + "Could not write clusters");
+    }
+}
+
+ClusterVector<int32_t> ClusterFile::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);
+    }
+}
+
+ClusterVector<int32_t>
+ClusterFile::read_clusters_without_cut(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)) {
+            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()),
+                                  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);
+    if (m_gain_map)
+        clusters.apply_gain_map(m_gain_map->view());
+    return clusters;
+}
+
+ClusterVector<int32_t> ClusterFile::read_clusters_with_cut(size_t n_clusters) {
+    ClusterVector<int32_t> clusters(3, 3);
+    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) {
+            Cluster3x3 c = read_one_cluster();
+            if (is_selected(c)) {
+                clusters.push_back(c.x, c.y,
+                                   reinterpret_cast<std::byte *>(c.data));
+            }
+        }
+    }
+
+    // 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) {
+                    Cluster3x3 c = read_one_cluster();
+                    if (is_selected(c)) {
+                        clusters.push_back(
+                            c.x, c.y, reinterpret_cast<std::byte *>(c.data));
+                    }
+                }
+            }
+
+            // we have enough clusters, break out of the outer while loop
+            if (clusters.size() >= n_clusters)
+                break;
+        }
+    }
+    if (m_gain_map)
+        clusters.apply_gain_map(m_gain_map->view());
+
+    return clusters;
+}
+
+Cluster3x3 ClusterFile::read_one_cluster() {
+    Cluster3x3 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;
+}
+
+ClusterVector<int32_t> ClusterFile::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();
+    }
+}
+
+ClusterVector<int32_t> ClusterFile::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<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(LOCATION + "Could not read clusters");
+    }
+    clusters.resize(n_clusters);
+    if (m_gain_map)
+        clusters.apply_gain_map(m_gain_map->view());
+    return clusters;
+}
+
+ClusterVector<int32_t> ClusterFile::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<int32_t> clusters(3, 3);
+    clusters.reserve(m_num_left);
+    clusters.set_frame_number(frame_number);
+    while (m_num_left) {
+        Cluster3x3 c = read_one_cluster();
+        if (is_selected(c)) {
+            clusters.push_back(c.x, c.y, reinterpret_cast<std::byte *>(c.data));
+        }
+    }
+    if (m_gain_map)
+        clusters.apply_gain_map(m_gain_map->view());
+    return clusters;
+}
+
+bool ClusterFile::is_selected(Cluster3x3 &cl) {
+    // Should fail fast
+    if (m_roi) {
+        if (!(m_roi->contains(cl.x, cl.y))) {
+            return false;
+        }
+    }
+    if (m_noise_map) {
+        int32_t sum_1x1 = cl.data[4];   // central pixel
+        int32_t sum_2x2 = cl.sum_2x2(); // highest sum of 2x2 subclusters
+        int32_t sum_3x3 = 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 || sum_3x3 <= 3 * noise) {
+            return false;
+        }
+    }
+    // we passed all checks
+    return true;
+}
+
+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});
+
+    if (clusters.cluster_size_x() == 3 || clusters.cluster_size_y() == 3) {
+        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;
+        }
+    } else if (clusters.cluster_size_x() == 2 ||
+               clusters.cluster_size_y() == 2) {
+        for (size_t i = 0; i < clusters.size(); i++) {
+            auto e = calculate_eta2(clusters.at<Cluster2x2>(i));
+            eta2(i, 0) = e.x;
+            eta2(i, 1) = e.y;
+        }
+    } else {
+        throw std::runtime_error("Only 3x3 and 2x2 clusters are supported");
+    }
+
+    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();
+    eta.sum = tot2[c];
+    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;
+}
+
+Eta2 calculate_eta2(Cluster2x2 &cl) {
+    Eta2 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.data[0] + cl.data[1] + cl.data[2] + cl.data[3];
+    eta.c = cBottomLeft; // TODO! This is not correct, but need to put something
+    return eta;
+}
+
+} // namespace aare

src/ClusterFile.test.cpp

@@ -10,7 +10,7 @@ using aare::Cluster;
 using aare::ClusterFile;
 using aare::ClusterVector;
 
-TEST_CASE("Read one frame from a cluster file", "[.files]") {
+TEST_CASE("Read one frame from a cluster file", "[.with-data]") {
     // We know that the frame has 97 clusters
     auto fpath = test_data_path() / "clust" / "single_frame_97_clustrers.clust";
     REQUIRE(std::filesystem::exists(fpath));
@@ -19,15 +19,14 @@ TEST_CASE("Read one frame from a cluster file", "[.files]") {
     auto clusters = f.read_frame();
     CHECK(clusters.size() == 97);
     CHECK(clusters.frame_number() == 135);
-    CHECK(clusters.at(0).x == 1);
+    CHECK(clusters[0].x == 1);
-    CHECK(clusters.at(0).y == 200);
+    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.at(0).data),
+    CHECK(std::equal(std::begin(clusters[0].data), std::end(clusters[0].data),
-                     std::end(clusters.at(0).data),
                      std::begin(expected_cluster_data)));
 }
 
-TEST_CASE("Read one frame using ROI", "[.files]") {
+TEST_CASE("Read one frame using ROI", "[.with-data]") {
     // We know that the frame has 97 clusters
     auto fpath = test_data_path() / "clust" / "single_frame_97_clustrers.clust";
     REQUIRE(std::filesystem::exists(fpath));
@@ -45,22 +44,21 @@ TEST_CASE("Read one frame using ROI", "[.files]") {
 
     // Check that all clusters are within the ROI
     for (size_t i = 0; i < clusters.size(); i++) {
-        auto c = clusters.at(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.at(0).x == 1);
+    CHECK(clusters[0].x == 1);
-    CHECK(clusters.at(0).y == 200);
+    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.at(0).data),
+    CHECK(std::equal(std::begin(clusters[0].data), std::end(clusters[0].data),
-                     std::end(clusters.at(0).data),
                      std::begin(expected_cluster_data)));
 }
 
-TEST_CASE("Read clusters from single frame file", "[.files]") {
+TEST_CASE("Read clusters from single frame file", "[.with-data]") {
 
     //  frame_number, num_clusters   [135] 97
     // [  1 200] [0 1 2 3 4 5 6 7 8]
@@ -162,6 +160,7 @@ TEST_CASE("Read clusters from single frame file", "[.files]") {
     // [ 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") {
@@ -170,10 +169,10 @@ TEST_CASE("Read clusters from single frame file", "[.files]") {
         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.at(0).x == 1);
+        REQUIRE(clusters[0].x == 1);
-        REQUIRE(clusters.at(0).y == 200);
+        REQUIRE(clusters[0].y == 200);
-        CHECK(std::equal(std::begin(clusters.at(0).data),
+        CHECK(std::equal(std::begin(clusters[0].data),
-                         std::end(clusters.at(0).data),
+                         std::end(clusters[0].data),
                          std::begin(expected_cluster_data)));
     }
     SECTION("Read more clusters than available") {
@@ -183,10 +182,10 @@ TEST_CASE("Read clusters from single frame file", "[.files]") {
         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.at(0).x == 1);
+        REQUIRE(clusters[0].x == 1);
-        REQUIRE(clusters.at(0).y == 200);
+        REQUIRE(clusters[0].y == 200);
-        CHECK(std::equal(std::begin(clusters.at(0).data),
+        CHECK(std::equal(std::begin(clusters[0].data),
-                         std::end(clusters.at(0).data),
+                         std::end(clusters[0].data),
                          std::begin(expected_cluster_data)));
     }
     SECTION("Read all clusters") {
@@ -194,16 +193,16 @@ TEST_CASE("Read clusters from single frame file", "[.files]") {
         auto clusters = f.read_clusters(97);
         REQUIRE(clusters.size() == 97);
         REQUIRE(clusters.frame_number() == 135);
-        REQUIRE(clusters.at(0).x == 1);
+        REQUIRE(clusters[0].x == 1);
-        REQUIRE(clusters.at(0).y == 200);
+        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.at(0).data),
+        CHECK(std::equal(std::begin(clusters[0].data),
-                         std::end(clusters.at(0).data),
+                         std::end(clusters[0].data),
                          std::begin(expected_cluster_data)));
     }
 }
 
-TEST_CASE("Read clusters from single frame file with ROI", "[.files]") {
+TEST_CASE("Read clusters from single frame file with ROI", "[.with-data]") {
     auto fpath = test_data_path() / "clust" / "single_frame_97_clustrers.clust";
     REQUIRE(std::filesystem::exists(fpath));
 
@@ -220,15 +219,14 @@ TEST_CASE("Read clusters from single frame file with ROI", "[.files]") {
 
     CHECK(clusters.size() == 10);
     CHECK(clusters.frame_number() == 135);
-    CHECK(clusters.at(0).x == 1);
+    CHECK(clusters[0].x == 1);
-    CHECK(clusters.at(0).y == 200);
+    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.at(0).data),
+    CHECK(std::equal(std::begin(clusters[0].data), std::end(clusters[0].data),
-                     std::end(clusters.at(0).data),
                      std::begin(expected_cluster_data)));
 }
 
-TEST_CASE("Read cluster from multiple frame file", "[.files]") {
+TEST_CASE("Read cluster from multiple frame file", "[.with-data]") {
 
     using ClusterType = Cluster<double, 2, 2>;
 
@@ -281,7 +279,7 @@ TEST_CASE("Read cluster from multiple frame file", "[.files]") {
     }
 }
 
-TEST_CASE("Write cluster with potential padding", "[.files][.ClusterFile]") {
+TEST_CASE("Write cluster with potential padding", "[.with-data][.ClusterFile]") {
 
     using ClusterType = Cluster<double, 3, 3>;
 
@@ -309,24 +307,24 @@ TEST_CASE("Write cluster with potential padding", "[.files][.ClusterFile]") {
     CHECK(read_cluster_vector.size() == 2);
     CHECK(read_cluster_vector.frame_number() == 0);
 
-    CHECK(read_cluster_vector.at(0).x == clustervec.at(0).x);
+    CHECK(read_cluster_vector[0].x == clustervec[0].x);
-    CHECK(read_cluster_vector.at(0).y == clustervec.at(0).y);
+    CHECK(read_cluster_vector[0].y == clustervec[0].y);
-    CHECK(std::equal(clustervec.at(0).data, clustervec.at(0).data + 9,
+    CHECK(std::equal(
-                     read_cluster_vector.at(0).data, [](double a, double b) {
+        clustervec[0].data.begin(), clustervec[0].data.end(),
-                         return std::abs(a - b) <
+        read_cluster_vector[0].data.begin(), [](double a, double b) {
-                                std::numeric_limits<double>::epsilon();
+            return std::abs(a - b) < std::numeric_limits<double>::epsilon();
         }));
 
-    CHECK(read_cluster_vector.at(1).x == clustervec.at(1).x);
+    CHECK(read_cluster_vector[1].x == clustervec[1].x);
-    CHECK(read_cluster_vector.at(1).y == clustervec.at(1).y);
+    CHECK(read_cluster_vector[1].y == clustervec[1].y);
-    CHECK(std::equal(clustervec.at(1).data, std::end(clustervec.at(1).data),
+    CHECK(std::equal(
-                     read_cluster_vector.at(1).data, [](double a, double b) {
+        clustervec[1].data.begin(), clustervec[1].data.end(),
-                         return std::abs(a - b) <
+        read_cluster_vector[1].data.begin(), [](double a, double b) {
-                                std::numeric_limits<double>::epsilon();
+            return std::abs(a - b) < std::numeric_limits<double>::epsilon();
         }));
 }
 
-TEST_CASE("Read frame and modify cluster data", "[.files][.ClusterFile]") {
+TEST_CASE("Read frame and modify cluster data", "[.with-data][.ClusterFile]") {
     auto fpath = test_data_path() / "clust" / "single_frame_97_clustrers.clust";
     REQUIRE(std::filesystem::exists(fpath));
 
@@ -341,10 +339,9 @@ TEST_CASE("Read frame and modify cluster data", "[.files][.ClusterFile]") {
         Cluster<int32_t, 3, 3>{0, 0, {0, 1, 2, 3, 4, 5, 6, 7, 8}});
 
     CHECK(clusters.size() == 98);
-    CHECK(clusters.at(0).x == 1);
+    CHECK(clusters[0].x == 1);
-    CHECK(clusters.at(0).y == 200);
+    CHECK(clusters[0].y == 200);
 
-    CHECK(std::equal(std::begin(clusters.at(0).data),
+    CHECK(std::equal(std::begin(clusters[0].data), std::end(clusters[0].data),
-                     std::end(clusters.at(0).data),
                      std::begin(expected_cluster_data)));
 }