A _process.h5 was internally inconsistent whenever the space group re-seated the lattice. The per-image reflections and lattices go to file as each image is processed, but the conventional setting is only chosen afterwards, so the file kept pre-reindex indices beside a post-reindex cell. Measured on an affected dataset, max|d_file - d(hkl, stored cell)| was 18.6 A. --mode scale then compared each frame's lattice against the stored cell, found none of 1800 acceptable, rejected every observation and died in the merge with "resolution calculation failed". /entry/MX/reindexMatrix now carries M with hkl_cell = M . hkl_written, and the reader applies it, so everything it hands out is in the setting of /entry/sample/unit_cell. Absent means the identity, so a file written before this reads exactly as before. On the affected dataset M comes out [[1,1,0],[0,1,1],[1,0,1]], det 2 - the primitive-to-body-centred basis its volume ratio implied - and the same measure falls from 18.6 A to 1.8e-5. Writing the reflections in the final setting instead was rejected: the per-image writer is shared with the broker, which streams and cannot buffer a run; and h,k,l, predicted_x/y and the per-image lattice are one consistent statement about one image, which retro-editing the indices would silently break. Two things turned up while fixing it. There are three re-seat sites, not one - the space-group search's own centred-lattice test re-seats too, and logged nothing - so the matrix composes over all of them. And the space group itself was never written: it was set only on the arm that searches, while a two-pass rotation run reuses pass 1's group and takes the other arm, so the canonical file carried a cell but no group and --mode scale merged in P1. --mode scale now reproduces --mode mx on the affected dataset: same space group and cell, 39329 unique reflections both, 100% of reflections common, CC 0.99985, sum|dI|/sum|I| = 0.0081. The residual is three reflections in 11.09 M crossing an ice-band edge, because mx carries the integrator's d and scale recomputes it from the cell. An unaffected dataset is byte-identical in .mtz, .hkl and .cif. An older affected file still cannot merge - M is not recoverable from it - but now says so in 51 s, naming both cells and the -S/-C override to use, instead of failing inside the merge. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CHMmeM1d489zvNFT7ZMN2P
437 lines
15 KiB
C++
437 lines
15 KiB
C++
// SPDX-FileCopyrightText: 2025 Filip Leonarski, Paul Scherrer Institute <filip.leonarski@psi.ch>
|
|
// SPDX-License-Identifier: GPL-3.0-only
|
|
|
|
#pragma once
|
|
|
|
#include <string>
|
|
#include <array>
|
|
#include <cstdint>
|
|
#include <map>
|
|
#include <vector>
|
|
#include <optional>
|
|
#include <cstring>
|
|
#include <stdexcept>
|
|
#include <nlohmann/json.hpp>
|
|
|
|
#include "CompressedImage.h"
|
|
#include "SpotToSave.h"
|
|
#include "UnitCell.h"
|
|
#include "GoniometerAxis.h"
|
|
#include "DetectorTransformation.h"
|
|
#include "GridScanSettings.h"
|
|
#include "Reflection.h"
|
|
#include "CrystalLattice.h"
|
|
#include "IndexingSettings.h"
|
|
#include "XrayFluorescenceSpectrum.h"
|
|
#include "../gemmi_gph/gemmi/symmetry.hpp"
|
|
|
|
constexpr const uint64_t user_data_release = 6;
|
|
constexpr const uint64_t user_data_magic_number = 0x52320000UL | user_data_release;
|
|
|
|
enum class CBORImageType {START, END, IMAGE, CALIBRATION, METADATA, NONE};
|
|
|
|
enum class FileWriterFormat : int {
|
|
DataOnly = 0,
|
|
NXmxLegacy = 1,
|
|
NXmxVDS = 2,
|
|
NXmxIntegrated = 3,
|
|
// 4 (CBF) and 5 (TIFF) removed - only HDF5 is written now. The values are kept
|
|
// (deprecated) in the OpenAPI enum in broker/jfjoch_api.yaml for back compatibility.
|
|
NoFile = 6
|
|
};
|
|
|
|
struct ROIConfigBox {
|
|
int64_t xmin, xmax, ymin, ymax;
|
|
};
|
|
|
|
struct ROIConfigCircle {
|
|
double r, x, y;
|
|
};
|
|
|
|
struct ROIConfigAzim {
|
|
double qmin, qmax;
|
|
// Azimuthal-angle sector in degrees; phi_min == phi_max means full ring.
|
|
float phi_min, phi_max;
|
|
};
|
|
|
|
struct ROIConfig {
|
|
enum class ROIType {Box, Circle, Azim} type;
|
|
std::string name;
|
|
union {
|
|
ROIConfigBox box;
|
|
ROIConfigCircle circle;
|
|
ROIConfigAzim azim;
|
|
};
|
|
};
|
|
|
|
struct ROIMessage {
|
|
int64_t sum;
|
|
uint64_t sum_square;
|
|
int64_t max_count;
|
|
uint64_t pixels;
|
|
uint64_t pixels_masked; // only used in the viewer for now
|
|
int64_t x_weighted;
|
|
int64_t y_weighted;
|
|
};
|
|
|
|
struct LatticeMessage {
|
|
char centering;
|
|
int64_t niggli_class;
|
|
gemmi::CrystalSystem crystal_system;
|
|
};
|
|
|
|
struct SmargonPosition {
|
|
float phi_deg = 0;
|
|
float chi_deg = 0;
|
|
Coord chi_axis = {0, 0, 1};
|
|
Coord phi_axis = {1, 0, 0}; // Default omega axis
|
|
};
|
|
|
|
struct DataMessage {
|
|
int64_t number = INT64_MIN;
|
|
CompressedImage image;
|
|
|
|
std::optional<uint64_t> packets_expected;
|
|
std::optional<uint64_t> packets_received;
|
|
std::optional<float> image_collection_efficiency;
|
|
|
|
std::optional<float> processing_time_s;
|
|
std::optional<float> spot_finding_time_s;
|
|
std::optional<float> azint_time_s;
|
|
std::optional<float> indexing_time_s;
|
|
std::optional<float> refinement_time_s;
|
|
std::optional<float> bragg_prediction_time_s;
|
|
std::optional<float> integration_time_s;
|
|
std::optional<float> preprocessing_time_s;
|
|
std::optional<float> compression_time_s;
|
|
std::optional<float> index_analysis_time_s;
|
|
std::optional<float> image_scale_time_s;
|
|
|
|
std::vector<SpotToSave> spots;
|
|
std::optional<int64_t> spot_count;
|
|
std::optional<int64_t> spot_count_ice_rings;
|
|
// Spots in the ICE-FREE control flanks either side of the hexagonal rings, rescaled to the ring
|
|
// bands' own q width - so spot_count_ice_rings / spot_count_ice_control is 1 when spots are spread
|
|
// evenly and > 1 when they pile up on the rings. Pooled over the run (a ratio of two per-image
|
|
// counts this small is meaningless on its own image), it is the second ice channel: TEXTURED ice
|
|
// arrives as discrete spots and leaves the radial profile - and so ice_ring_score - flat.
|
|
std::optional<float> spot_count_ice_control;
|
|
std::optional<int64_t> spot_count_low_res;
|
|
|
|
std::vector<float> spot_plot_count;
|
|
std::vector<float> spot_plot_intensity;
|
|
std::vector<float> spot_plot_one_over_d_square;
|
|
|
|
std::vector<float> az_int_profile;
|
|
std::vector<float> az_int_profile_std;
|
|
std::vector<uint64_t> az_int_profile_count;
|
|
|
|
std::optional<float> bkg_estimate;
|
|
std::optional<float> ice_ring_score; // strongest ice ring over the smooth radial background (1 = none)
|
|
|
|
std::optional<bool> indexing_result;
|
|
std::optional<CrystalLattice> indexing_lattice;
|
|
std::vector<CrystalLattice> indexing_extra_lattices;
|
|
std::optional<int64_t> indexing_lattice_count;
|
|
std::optional<UnitCell> indexing_unit_cell;
|
|
std::optional<int64_t> spot_count_indexed;
|
|
|
|
std::optional<float> profile_radius;
|
|
std::optional<float> mosaicity_deg;
|
|
|
|
std::optional<float> b_factor;
|
|
|
|
std::vector<uint64_t> adu_histogram;
|
|
|
|
std::optional<int64_t> integrated_reflections;
|
|
|
|
uint64_t timestamp;
|
|
uint32_t timestamp_base;
|
|
|
|
uint32_t exptime;
|
|
uint32_t exptime_base;
|
|
|
|
std::string run_name;
|
|
uint64_t run_number;
|
|
|
|
std::optional<int64_t> pixel_sum;
|
|
std::optional<int64_t> saturated_pixel_count;
|
|
std::optional<int64_t> error_pixel_count;
|
|
std::optional<int64_t> strong_pixel_count;
|
|
std::optional<int64_t> min_viable_pixel_value;
|
|
std::optional<int64_t> max_viable_pixel_value;
|
|
std::optional<float> resolution_estimate;
|
|
|
|
nlohmann::json user_data;
|
|
|
|
std::optional<uint64_t> jf_info;
|
|
std::optional<uint64_t> receiver_aq_dev_delay;
|
|
std::optional<uint64_t> receiver_buf_available;
|
|
std::optional<uint64_t> receiver_buf_in_sending;
|
|
std::optional<uint64_t> receiver_buf_in_preparation;
|
|
std::optional<uint64_t> storage_cell;
|
|
|
|
std::optional<uint64_t> xfel_pulse_id;
|
|
std::optional<uint64_t> xfel_event_code;
|
|
|
|
std::map<std::string, ROIMessage> roi;
|
|
|
|
std::optional<int64_t> original_number;
|
|
std::vector<Reflection> reflections;
|
|
|
|
std::vector<float> integration_B_logI;
|
|
std::vector<float> integration_B_one_over_d_square;
|
|
|
|
std::vector<float> integration_Isigma;
|
|
std::vector<float> integration_Isigma_one_over_d_square;
|
|
|
|
std::optional<float> beam_corr_x;
|
|
std::optional<float> beam_corr_y;
|
|
|
|
std::optional<LatticeMessage> lattice_type;
|
|
|
|
std::optional<float> image_scale_factor;
|
|
std::optional<float> image_scale_cc;
|
|
std::optional<float> image_scale_mosaicity;
|
|
|
|
std::optional<float> compression_ratio;
|
|
};
|
|
|
|
struct HDF5DataSourceMessage {
|
|
std::string filename;
|
|
std::string dataset = "/entry/data/data";
|
|
|
|
uint64_t source_first_image = 0;
|
|
uint64_t virtual_first_image = 0;
|
|
uint64_t image_count = 0;
|
|
};
|
|
|
|
struct StartMessage {
|
|
float detector_distance;
|
|
float beam_center_x;
|
|
float beam_center_y;
|
|
|
|
uint64_t number_of_images;
|
|
|
|
std::optional<uint64_t> images_per_trigger;
|
|
|
|
uint64_t image_size_x;
|
|
uint64_t image_size_y;
|
|
// Whether the assembled image is mirrored in Y relative to the detector's raw readout. True is
|
|
// the MX convention - row 0 at the top of the detector, seen from the sample - and is what
|
|
// Jungfraujoch has always produced, so it is also what absence of the field means.
|
|
bool mirror_y = true;
|
|
uint64_t bit_depth_image; // user data
|
|
std::optional<uint64_t> bit_depth_readout;
|
|
bool pixel_signed; // user data
|
|
|
|
bool countrate_correction_enabled;
|
|
|
|
float incident_energy;
|
|
float incident_wavelength;
|
|
std::optional<float> incident_wavelength_spread; // NXmx incident_wavelength_spread: FWHM of dlambda (Angstrom)
|
|
|
|
float frame_time;
|
|
float count_time;
|
|
|
|
int64_t saturation_value;
|
|
std::optional<int64_t> error_value;
|
|
std::optional<int64_t> underload_value; // NXmx: lowest valid value
|
|
|
|
float pixel_size_x;
|
|
float pixel_size_y;
|
|
float sensor_thickness;
|
|
std::string sensor_material;
|
|
|
|
std::optional<UnitCell> unit_cell; // user data
|
|
std::optional<uint64_t> space_group_number; // user data
|
|
uint64_t max_spot_count; // user data
|
|
uint64_t max_extra_lattices = 0;
|
|
|
|
std::optional<uint64_t> storage_cell_number;
|
|
uint64_t storage_cell_delay_ns;
|
|
|
|
bool flatfield_enabled;
|
|
bool pixel_mask_enabled;
|
|
|
|
std::string arm_date;
|
|
|
|
std::string sample_name; // user data
|
|
std::string file_prefix; // user data
|
|
int64_t images_per_file = 1; // user data
|
|
|
|
std::vector<std::string> channels;
|
|
|
|
std::string detector_description;
|
|
std::string detector_serial_number;
|
|
std::string run_name;
|
|
uint64_t run_number;
|
|
|
|
std::vector<std::string> gain_file_names;
|
|
|
|
std::vector<ROIConfig> rois;
|
|
|
|
std::optional<GridScanSettings> grid_scan;
|
|
std::optional<GoniometerAxis> goniometer;
|
|
float detector_translation[3];
|
|
|
|
std::string source_type;
|
|
std::string source_name;
|
|
std::string instrument_name;
|
|
|
|
uint64_t summation;
|
|
|
|
std::vector<float> az_int_bin_to_q;
|
|
std::vector<float> az_int_bin_to_two_theta;
|
|
std::vector<float> az_int_bin_to_phi;
|
|
std::optional<uint64_t> az_int_q_bin_count;
|
|
std::optional<uint64_t> az_int_phi_bin_count;
|
|
std::vector<uint16_t> az_int_map;
|
|
|
|
std::map<std::string, std::vector<uint32_t>> pixel_mask;
|
|
|
|
// Per-pixel ROI bitmask (converted geometry), bit i set for ROI i in rois.
|
|
std::vector<uint16_t> roi_map;
|
|
|
|
std::map<std::string, float> threshold_energy;
|
|
|
|
std::optional<float> total_flux;
|
|
std::optional<float> attenuator_transmission;
|
|
|
|
std::optional<bool> write_master_file;
|
|
std::optional<bool> write_images;
|
|
|
|
nlohmann::json user_data;
|
|
|
|
std::optional<float> data_reduction_factor_serialmx;
|
|
std::string experiment_group;
|
|
|
|
std::string jfjoch_release;
|
|
|
|
std::optional<uint64_t> socket_number; // This is number of socket in ZeroMQ
|
|
std::string writer_notification_zmq_addr; // Socket to inform detector on writer done
|
|
|
|
std::optional<bool> jungfrau_conversion_enabled;
|
|
std::optional<float> jungfrau_conversion_factor;
|
|
std::optional<bool> geometry_transformation_enabled;
|
|
std::optional<std::string> summation_mode;
|
|
std::optional<bool> overwrite;
|
|
std::optional<FileWriterFormat> file_format;
|
|
std::optional<bool> xfel_pulse_id;
|
|
std::optional<float> ring_current_mA;
|
|
std::optional<float> sample_temperature_K;
|
|
|
|
IndexingAlgorithmEnum indexing_algorithm;
|
|
GeomRefinementAlgorithmEnum geom_refinement_algorithm;
|
|
|
|
std::optional<float> poni_rot1;
|
|
std::optional<float> poni_rot2;
|
|
std::optional<float> poni_rot3;
|
|
|
|
XrayFluorescenceSpectrum fluorescence_spectrum;
|
|
|
|
std::optional<bool> detect_ice_rings;
|
|
|
|
std::vector<HDF5DataSourceMessage> hdf5_source_data;
|
|
std::optional<std::string> master_suffix;
|
|
|
|
std::optional<SmargonPosition> smargon_position;
|
|
};
|
|
|
|
struct EndMessage {
|
|
// The sample transformation chain, base first (see DetectorTransformation). Optional: when it is
|
|
// empty the writer builds the same chain itself from the start message, so a producer that does
|
|
// not send it loses nothing and the wire format needs no version bump. It lives in the END
|
|
// message because a future producer will want to report where the stage actually WENT, and that
|
|
// is only known once the run is over.
|
|
std::vector<DetectorTransformation> transformations;
|
|
uint64_t max_image_number; // Counting from 1, i.e. 0 = no images collected
|
|
std::optional<uint64_t> images_collected_count;
|
|
std::optional<uint64_t> images_sent_to_write_count;
|
|
std::optional<uint64_t> max_receiver_delay;
|
|
std::optional<float> efficiency;
|
|
std::optional<float> indexing_rate;
|
|
std::optional<float> bkg_estimate;
|
|
|
|
std::optional<std::string> end_date;
|
|
|
|
std::string run_name;
|
|
uint64_t run_number;
|
|
|
|
std::map<std::string, std::vector<float>> az_int_result;
|
|
std::map<std::string, std::vector<uint64_t>> adu_histogram;
|
|
uint64_t adu_histogram_bin_width;
|
|
|
|
std::optional<LatticeMessage> rotation_lattice_type;
|
|
std::optional<CrystalLattice> rotation_lattice;
|
|
std::vector<CrystalLattice> rotation_extra_lattices;
|
|
std::optional<UnitCell> unit_cell;
|
|
// Space group determined by the offline analysis (overrides the start message when writing the
|
|
// master, since it is only known after merging).
|
|
std::optional<uint64_t> space_group_number;
|
|
// Change of basis (3x3 integers, row major) from the setting the PER-IMAGE reflections and
|
|
// lattices were written in to the setting of unit_cell / rotation_lattice above:
|
|
// hkl_cell = M . hkl_written, and the same M takes each per-image lattice across. The per-image
|
|
// data go to file as each image is processed, but the space group - and with it the conventional
|
|
// setting the merge works in - is only chosen afterwards, so the two settings are not always the
|
|
// same one. Absent means they are, i.e. the identity.
|
|
std::optional<std::array<int32_t, 9>> reindex_matrix;
|
|
|
|
// Geometry refined by the offline analysis. When present these overwrite the master-file geometry
|
|
// at finalization (rugnux writes the refined values; the broker leaves them empty and keeps the
|
|
// user-provided StartMessage geometry). Beam centre in pixels, detector tilt (PONI rot1/2/3) in
|
|
// radians, rotation axis as a unit vector.
|
|
std::optional<float> refined_beam_center_x;
|
|
std::optional<float> refined_beam_center_y;
|
|
std::optional<float> refined_poni_rot1;
|
|
std::optional<float> refined_poni_rot2;
|
|
std::optional<float> refined_poni_rot3;
|
|
std::optional<Coord> refined_rotation_axis;
|
|
|
|
// Vectors with end result:
|
|
std::vector<float> data_collection_efficiency;
|
|
std::vector<int32_t> spot_count;
|
|
std::vector<int32_t> spot_count_ice_ring;
|
|
std::vector<float> spot_count_ice_control;
|
|
std::vector<int32_t> spot_count_low_res;
|
|
std::vector<int32_t> spot_count_indexed;
|
|
std::vector<uint8_t> image_indexed;
|
|
std::vector<int32_t> indexed_lattice_count;
|
|
std::vector<float> v_bkg_estimate;
|
|
std::vector<float> profile_radius;
|
|
std::vector<float> mosaicity;
|
|
std::vector<float> bFactor;
|
|
std::vector<float> resolution_estimate;
|
|
std::vector<int64_t> min_viable_pixel_value;
|
|
std::vector<int64_t> max_viable_pixel_value;
|
|
std::vector<int32_t> saturated_pixel_count;
|
|
std::vector<int32_t> error_pixel_count;
|
|
std::vector<int32_t> integrated_reflections;
|
|
std::vector<uint8_t> niggli_class;
|
|
std::vector<int64_t> pixel_sum;
|
|
|
|
std::vector<float> image_scale_factor;
|
|
std::vector<float> image_scale_cc;
|
|
std::vector<float> image_scale_mosaicity;
|
|
// Per-image ice strength. Note the name carries no v_ prefix, unlike v_bkg_estimate above - it
|
|
// was shipped that way and the CBOR key is part of the stream format.
|
|
std::vector<float> ice_ring_score;
|
|
// Run mean of the above, the single "how icy was this dataset" number (1 = no ice). The
|
|
// bkg_estimate scalar's counterpart; written to /entry/MX/iceRingScoreMean.
|
|
std::optional<float> ice_ring_score_mean;
|
|
|
|
// Per-image sweep-quality code: 0 = the image falls in no flagged range, otherwise 1 + the
|
|
// SweepQualityReason of the range it falls in (image_analysis/scale_merge/Merge.h). Written to
|
|
// /entry/MX/sweepQuality, with the vocabulary alongside it in /entry/MX/sweepQualityReasons.
|
|
// Empty when the sweep-quality diagnostic did not run, so nothing could be looked for - which is
|
|
// not the same as every image being clean. Offline (rugnux) only; the broker does not merge.
|
|
std::vector<uint8_t> sweep_quality;
|
|
std::vector<std::string> sweep_quality_reasons;
|
|
};
|
|
|
|
struct MetadataMessage {
|
|
std::string run_name;
|
|
uint64_t run_number;
|
|
std::vector<DataMessage> images;
|
|
};
|