Files
Jungfraujoch/common/ScalingSettings.h
T
leonarski_fandClaude Opus 5 16bf3408f0 Address code-review findings; make detection limits detector-driven
One changeset, developed together in response to a review of this branch, so the
files carry several of the changes at once. Full test suite passes (733 cases).

Spot finding
- Split ImageSpotFinder into Detect() (flag strong pixels - the expensive
  per-pixel pass) and ExtractSpots() (CCL + min/max-pix + resolution mask), with
  Run() = both. The per-image min-pix escalation now detects ONCE and repeats
  only the cheap extraction, instead of re-running the whole finder four times
  per frame as it did on the default path. It also keeps the winning attempt's
  spot list rather than re-extracting it, so the frame that is integrated is
  exactly the frame that was scored - which a GPU re-extract could not guarantee
  (float atomic ordering).
- spot_finding_time_s no longer swallows indexing time, and indexing_time_s now
  sums every escalation call instead of reporting only the last.

Detection limits follow the detector
- The azimuthal-integration upper q and the spot-finding high-resolution limit
  are now std::optional, in the C++ structs AND in the OpenAPI schema, and
  resolve to the detector's own maximum (DiffractionExperiment::GetDetectorMaxQ_
  recipA). Adaptive detection reads a pixel's ring from the azimuthal bins, so a
  pixel outside that q range could never be strong - the integration range
  silently bounded what detection could see, regardless of the requested
  resolution limit. Regenerated the C++ and TypeScript clients; the viewer and
  the web frontend each gained a "to detector edge" switch.

Detection defaults are now per workflow (measured, not assumed)
- Stills: adaptive detection, min-pix chosen per image, no resolution clipping.
- Rotation: fixed-threshold finder, min-pix 2, 1.5 A limit.
  On a 33-crystal rotation battery, adaptive detection helped four hard crystals
  but deterministically broke three (a lost space group, a halved indexing rate,
  a collapsed merge), and the detector-edge limit cost indexing on a strong
  rotation set (100.0 -> 96.8%). Each is still overridable by its flag, and
  --no-adaptive-spots is new.

Indexer seed escalation
- Stop escalating once a seed's lattice explains >= 90% of the seed spots.
  Previously any frame with >= 80 spots always paid three indexer calls, online
  broker included.

Merge-consistency filter
- --min-image-cc gated on a per-image CC computed BEFORE the stills partiality
  post-refinement and never refreshed; the refiner now recomputes it, so the
  reported CC describes the data that are actually merged.
- Replaced the per-call cc_mask argument with one MergeOnTheFly flag, so the
  merge, the error model and MergeStats can no longer disagree about which
  images are in (the --scale path merged unfiltered while its statistics were
  filtered).

Per-image B-factor refinement (-B) removed
- Measured on four serial-stills datasets: it is a no-op where the per-image fit
  is well conditioned and actively harmful where it is not (CC1/2 -8.1, R_meas
  +23.2 on the weakest large-cell set, whose fits hit their [-50, 200] bounds on
  14-25% of images). It had also been silently DISCARDED since the partiality
  post-refinement landed - reported but not applied. Rather than fix and keep a
  knob with no demonstrated benefit, the flag and the whole image_scale_b_factor
  chain are gone: setting, scaling fit, message field, CBOR, HDF5 write and
  read-back, per-image plot, OpenAPI enum, viewer column and checkbox, docs.
  ScaleOnTheFly no longer needs Ceres at all - the fit is a linear IRLS.
  (The Wilson per-image b_factor is a different quantity and stays.)

Stills partiality width now fits both of its components
- sigma^2 = gamma0^2 + (gamma_e*d*)^2 instead of a purely angular gamma_e*d*
  with gamma0 pinned to 0. Fitted per crystal by least squares of dist_ewald^2
  on d*^2. The angular-only width is fitted over a d*^2-dense population, so it
  was pinned by the high-resolution edge and collapsed at low d*: median
  partiality 0.008 beyond 13 A for reflections that were plainly recorded, 55%
  of them under the merge's partiality floor, and the survivors divided by those
  values - which inflated the merged low-resolution intensity scale 3.6x
  (~ +9 A^2 of apparent B). Measured on 5000 stills: the ramp flattens to 0.89x,
  no observation is dropped any more (701750 -> 716811), shell-mean CC1/2 and
  R-free improve slightly. Note CC1/2, R_meas, completeness and a B-refining
  R-free are all blind to that ramp, which is why it survived earlier validation;
  the cost is high-resolution R_meas (98.5 -> 101.9 shell-averaged).

Removed dead code from add-then-remove churn
- Prediction-time "still partiality" (unreachable: no setter), the phantom
  IndexingSettings::min_indexed_spot_fraction knob (getter, no setter - now the
  constant it always was), StillsPartialityRefine's caller-less Settings
  constructor and its reference to a long-gone env var, ProcessImage's unread
  bool return, an unused include, and a dead viewer overlay hook.

Also
- Viewer: the magnifier compared a QImage with itself, so its scene rect was set
  once ever and it could not pan into a larger dataset; the hover tail timer
  could fire after leaveEvent and resurrect the resolution readout outside the
  image.
- update_version.sh regenerated the frontend lock file BEFORE bumping the
  version (every release shipped an off-by-one lock), and did git rm/git add on
  a path that has not existed since the client moved to src/client - with no
  set -e, both failed silently.
- fpga/pcie_driver/postinstall.sh tested "[ ! occurrences > 0 ]", which is a
  redirect, not a test, so dkms add never ran.
- Unit tests for the adaptive-threshold host functions, which had none.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-07-27 09:07:00 +02:00

146 lines
8.2 KiB
C++

// SPDX-FileCopyrightText: 2026 Filip Leonarski, Paul Scherrer Institute <filip.leonarski@psi.ch>
// SPDX-License-Identifier: GPL-3.0-only
#pragma once
#include <optional>
#include "JFJochException.h"
// How the high-resolution cutoff for the written reflections and the reported shell table is chosen
// when no explicit --scaling-high-resolution is given. Off = keep the full (detector-edge) range;
// CCHalfLogistic = fit the CC1/2 fall-off and cut one shell past cc_target (DIALS-style, generous).
enum class ResolutionCutoffMethod { Off, CCHalfLogistic };
class ScalingSettings {
bool refine_wedge = false;
bool merge_friedel = true;
std::optional<double> high_resolution_limit_A;
std::optional<double> wedge_for_scaling;
std::optional<double> forced_mosaicity; // diagnostic: fix the scaling mosaicity (deg) instead of the per-image seed
double min_partiality = 0.02;
// Capture-aware systematic uncertainty for the rot3d combine: a full reconstructed from only
// a fraction f<1 of its rocking curve is extrapolated, and the unobserved (1-f) carries a
// systematic error ~coeff*(1-f)*I that plain counting sigma misses. 0 = off (baseline).
double capture_uncertainty_coeff = 0.0;
// Full-level captured-fraction floor for the rot3d combine: drop a reconstructed full whose rocking
// curve was only fractionally captured (sum of its partials' partiality < this). Distinct from
// min_partiality, which gates individual partials; this gates the assembled full. 0 = off (baseline).
double min_captured_fraction = 0.0;
double min_cc_for_image = 0.0;
double outlier_reject_nsigma = 0.0; // per-observation merge outlier rejection (XDS/DIALS-style); 0 = off, e.g. 6 enables
// Scale fulls: after the rotation 3D combine, refit a per-frame scale on the combined fulls (XDS
// order). Only used by the rotation path (RotationScaleMerge).
bool scale_fulls = false;
// Correction surfaces fitted on the rot3d fulls after scale-fulls: a decay (per-run Debye-Waller B)
// and an absorption surface (over the diffracted-beam direction in the goniometer frame). Both are
// cross-validated, so they no-op when their systematic is absent - hence ON by default (they only ever
// help or do nothing). No-op without rot3d. Set false to disable both.
bool correction_surfaces = true;
// Absorption-surface refinement iteration count (used when correction_surfaces is on).
int absorption_iter = 3;
// Physical partiality post-refinement for the STILLS merge (StillsPartialityRefine): refine a per-crystal
// orientation tilt against the running merge, recompute each reflection's partiality from the refined
// geometry (angular Ewald-proximity model), and re-scale/merge - the "full model" for stills. ON by
// default (helps mono stills, neutral on pink beam, tames weak data via a soft prior). rugnux
// --simple-stills turns it OFF, reverting to treating every reflection as a full (p = 1, single pass).
bool stills_partiality_refine = true;
// Expected-variance merge weighting for the STILLS merge (MergeOnTheFly). When combining a reflection's
// redundant observations by inverse variance, rebuild the Poisson signal part of each observation's
// variance at the reflection's EXPECTED <I> instead of the observation's own intensity. Weighting by an
// observation's own sigma^2 biases the inverse-variance mean low at <1 photon (an up-fluctuated
// observation gets a larger sigma and is over-downweighted). Default on - it mirrors the rotation combine
// (RotationScaleMerge::process_rawrun), which already does this, and is R-free-neutral on strong data and
// better on weak. --no-expected-variance-merge restores the old observed-sigma weighting.
bool expected_variance_merge = true;
// Smooth the per-frame scale G across frames (centered moving average of log G) before the rot3d
// combine, so a rocking event's partials share a consistent scale. Given as a ROTATION RANGE in
// degrees (like XDS DELPHI), converted to an odd frame window from the oscillation step; this keeps
// the smoothing physical (independent of frame slicing). 0 = off. A no-op without rot3d.
double smooth_g_deg = 0.0;
// Per-batch relative-B on the rot3d fulls (beyond the single global decay slope): bin frames into
// rotation-range batches of this width in degrees and refine one relative Debye-Waller B per batch, so
// NON-monotonic changes in scattering power across the run (absorption path, crystal slippage, dose
// bursts) are corrected the resolution-flat per-frame G and the single decay slope both miss. Cross-
// validated + zero-mean-anchored, so a no-op when absent. 0 = off. A no-op without rot3d.
double relative_b_deg = 0.0;
double rfree_fraction = 0.05;
// Automatic high-resolution cutoff for the written reflections + reported shells (not the scaling
// or the error model, and not the per-image _process.h5). Applied only when no explicit
// high_resolution_limit_A is set - that manual limit always wins and disables the auto-cut.
ResolutionCutoffMethod resolution_cutoff = ResolutionCutoffMethod::CCHalfLogistic;
double resolution_cc_target = 0.30; // CC1/2 value defining the fall-off limit before the +1 shell
int report_shell_count = 10; // number of resolution shells in the reported statistics table
bool scaling_regularize = false;
public:
ScalingSettings& RefineRotationWedge(bool input);
ScalingSettings& RotationWedgeForScaling(std::optional<double> input);
ScalingSettings& MergeFriedel(bool input);
ScalingSettings& HighResolutionLimit_A(double limit);
ScalingSettings& HighResolutionLimit_A(std::optional<double> limit); // nullopt clears the limit
ScalingSettings& MinPartiality(double min_partiality);
ScalingSettings& ForcedMosaicity(std::optional<double> input);
ScalingSettings& CaptureUncertaintyCoeff(double input);
ScalingSettings& MinCapturedFraction(double input);
ScalingSettings& MinCCForImage(double min_cc_for_image);
ScalingSettings& OutlierRejectNsigma(double input);
ScalingSettings& ScaleFulls(bool input);
ScalingSettings& AbsorptionIter(int input);
ScalingSettings& CorrectionSurfaces(bool input);
ScalingSettings& StillsPartialityRefine(bool input);
ScalingSettings& ExpectedVarianceMerge(bool input);
ScalingSettings& SmoothGDegrees(double input);
ScalingSettings& RelativeBDegrees(double input);
ScalingSettings& RfreeFraction(double input);
ScalingSettings& ScalingRegularize(bool input);
ScalingSettings& ResolutionCutoff(ResolutionCutoffMethod input);
ScalingSettings& ResolutionCCTarget(double input);
ScalingSettings& ReportShellCount(int input);
[[nodiscard]] bool GetRefineWedge() const;
[[nodiscard]] double GetMinMosaicity() const;
[[nodiscard]] double GetDefaultMosaicity() const;
[[nodiscard]] double GetMaxMosaicity() const;
[[nodiscard]] double GetMinWedge() const;
[[nodiscard]] std::optional<double> GetRotationWedgeForScaling() const;
[[nodiscard]] double GetMaxWedge() const;
[[nodiscard]] bool GetMergeFriedel() const;
[[nodiscard]] std::optional<double> GetHighResolutionLimit_A() const;
[[nodiscard]] double GetMinPartiality() const;
[[nodiscard]] std::optional<double> GetForcedMosaicity() const;
[[nodiscard]] double GetCaptureUncertaintyCoeff() const;
[[nodiscard]] double GetMinCapturedFraction() const;
[[nodiscard]] double GetMinCCForImage() const;
[[nodiscard]] double GetOutlierRejectNsigma() const;
[[nodiscard]] bool GetScaleFulls() const;
[[nodiscard]] int GetAbsorptionIter() const;
[[nodiscard]] bool GetCorrectionSurfaces() const;
[[nodiscard]] bool GetStillsPartialityRefine() const;
[[nodiscard]] bool GetExpectedVarianceMerge() const;
[[nodiscard]] double GetSmoothGDegrees() const;
[[nodiscard]] double GetRelativeBDegrees() const;
[[nodiscard]] double GetRfreeFraction() const;
[[nodiscard]] bool GetScalingRegularize() const;
[[nodiscard]] ResolutionCutoffMethod GetResolutionCutoff() const;
[[nodiscard]] double GetResolutionCCTarget() const;
[[nodiscard]] int GetReportShellCount() const;
};