Files
Jungfraujoch/image_analysis/scale_merge/Merge.h
T
leonarski_fandClaude Opus 5 72efb75a8c Merging: do not floor the merged sigma at the systematic term
The merged sigma was floored at b*|I|, so I/sigma could never exceed the reported ISa.
On one dataset every merged reflection came out at I/sigma <= 12.96 with a 99th
percentile of 12.77 in every resolution shell alike, while the scatter of the
observations implied about 44 and XDS reported 58.

The floor is wrong in principle. `b` is fitted from the scatter BETWEEN a reflection's
symmetry equivalents, i.e. from the part that is not common to them, so it averages
down with multiplicity exactly like the counting term. 1/sqrt(sum_w) with the
b-inflated per-observation sigma already gives b*I/sqrt(n); flooring at b*|I| puts the
sqrt(n) back. That is the whole effect: 12.96 * sqrt(21.6) = 60, against XDS's 58.

It was introduced on a comparison of our MERGED I/sigma against XDS's UNMERGED
I/sigma. XDS's own merged low-resolution I/sigma exceeds its reported ISa on 30 of the
39 reference datasets here, median ratio 1.78 and up to 4.23.

Merged low-shell I/sigma now lands where XDS's does: 22.4 -> 46.2 against 46.2 on one
crystal, 26.7 -> 115.7 against 96.6 on another, 12.5 -> 45.0 against 58.0 on a third.
Over the 38-crystal battery the space groups, the merged reflection sets, R_meas and
CC1/2 are all unchanged - every one of them is sigma-independent, which is what makes
them the right control - and <I/sigma> rises on 35 crystals with none worse.

The asymptotic estimator that fed the floor stays, for the reported ISa only, and is
repaired in the process: it subtracts a*sigma^2 rather than the raw sigma^2 (at a < 1
the difference is the same size as the b^2 being measured, which is what made it
flip between 10.9 and 62.7 on consecutive passes of the same data), it rescales each
group's variance median-unbiased before subtracting an unbiased counting term, its
I/sigma gate uses the same convention, and it is bounded by the whole-range b - an
asymptote exists to refine 1/b upward, not to report 0.3 because "strong" was selected
on a sigma scale the fit itself rejects.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-09 20:58:35 +02:00

181 lines
8.7 KiB
C++

// SPDX-FileCopyrightText: 2025 Filip Leonarski, Paul Scherrer Institute <filip.leonarski@psi.ch>
// SPDX-License-Identifier: GPL-3.0-only
#pragma once
#include <algorithm>
#include <cmath>
#include <map>
#include <unordered_map>
#include <vector>
#include "../../common/Logger.h"
#include "../../common/DiffractionExperiment.h"
#include "../../common/Reflection.h"
#include "../IntegrationOutcome.h"
#include "HKLKey.h"
struct MergeStatisticsShell {
float d_min = 0.0f;
float d_max = 0.0f;
float mean_one_over_d2 = 0;
int total_observations = 0;
int unique_reflections = 0;
int possible_unique_reflections = 0;
double mean_i_over_sigma = 0.0;
double cc_half = 0.0f;
double cc_ref = NAN;
// Redundancy-independent merging R-factor (Diederichs & Karplus 1997), computed over the
// observations that enter the merge: R_meas = sum_hkl sqrt(n/(n-1)) sum_i|I_i-<I>| / sum I_i.
double r_meas = NAN;
// Anomalous signal-to-noise (XDS "SigAno" / mmCIF pdbx_absDiff_over_sigma_anomalous):
// <|I(+)-I(-)|> / <sigma(I(+)-I(-))> over acentric reflections measured in both hands. NaN when
// there is no anomalous split (e.g. Friedel-merged with no mates, or the stills path).
double abs_diff_over_sigma_anomalous = NAN;
};
struct MergeStatistics {
std::vector<MergeStatisticsShell> shells;
MergeStatisticsShell overall;
// Dataset-wide isotropic Wilson B-factor estimate (A^2) from the log-linear fit of the shell-mean
// merged intensity against 1/d^2 (CalcGlobalWilsonB) - the analogue of XDS's "WILSON LINE ... B=".
// Diagnostic only; not used in scaling. NaN when not determined.
double wilson_b = NAN;
double wilson_b_correlation = NAN;
// Radiation-damage monitor (rotation only): the relative Debye-Waller B change measured from the first
// to the last frame of the run (A^2; positive = high-resolution intensity fades with dose = damage) and
// the per-batch relative-B curve it was derived from. Measured before any decay/relative-B correction is
// applied, so it reports how much radiation damage was present. Diagnostic; NaN / empty for stills or
// when not determined. batch_deg is the rotation width per batch of the curve.
double radiation_damage_delta_b = NAN;
std::vector<float> radiation_damage_b_batch;
double radiation_damage_batch_deg = 0.0;
};
std::ostream &operator<<(std::ostream &output, const MergeStatisticsShell &in);
std::ostream &operator<<(std::ostream &output, const MergeStatistics &in);
struct MergeAccum {
int32_t h = 0;
int32_t k = 0;
int32_t l = 0;
float d = NAN;
double sum_wI = 0.0;
double sum_w = 0.0;
double sum_wI_half[2] = {0.0, 0.0};
double sum_w_half[2] = {0.0, 0.0};
size_t n_half[2] = {0, 0};
};
class MergeOnTheFly {
mutable std::mutex merged_mutex;
const int space_group_number = 1;
ScalingSettings scaling_settings;
IndexingSettings indexing_settings;
std::optional<UnitCell> reference_cell;
std::optional<double> high_resolution_limit;
std::optional<double> low_resolution_limit;
std::optional<double> image_cc_limit;
// Apply image_cc_limit in Mask(). One flag for the whole engine, not a per-call argument, so the
// merge, the error model and MergeStats can never disagree about which images are in.
bool filter_by_image_cc = false;
double min_partiality = 0.02;
// When set, ice-ring-flagged reflections are left out of this merge. Used for the P1 pass whose
// merged intensities feed the space-group search and the error model - those model fits must not
// see the ice-contaminated intensities. The final in-symmetry merge keeps them (for completeness).
bool exclude_ice_rings = false;
HKLKeyGenerator generator;
std::map<uint64_t, MergeAccum> accumulator;
// Global error model (XDS form): sigma_corr^2 = a*sigma^2 + (b*<I>)^2. a rescales the
// (under-estimated) counting variance; the (b*<I>)^2 term adds the intensity-
// proportional systematic error that counting statistics miss, so strong reflections
// are no longer over-weighted. ISa = 1/b is the asymptotic I/sigma. Refined from the
// scatter of symmetry equivalents (RefineErrorModel); identity until then.
bool error_model_active = false;
double error_model_a = 1.0;
double error_model_b = 0.0;
double error_model_chi2 = 0.0; // achieved median reduced chi^2 (~1.0 = honestly calibrated sigmas)
// The (b*I)^2 term uses the reflection's *mean* intensity (constant over its
// observations), so it inflates sigma without biasing the inverse-variance weights -
// using the per-observation I_i instead would over-weight down-fluctuated points.
std::unordered_map<uint64_t, float> error_model_mean_I;
[[nodiscard]] float CorrectedSigma(float I_corr, float sigma_corr, float image_scale_corr,
float var_bkg,
uint64_t hkl_key) const;
// Optional per-observation outlier rejection: drop observations whose corrected
// intensity lies more than reject_nsigma error-model sigmas from the reflection's
// *median* (a robust centre). The error-model sigma already captures the genuine
// (e.g. partiality) scatter, so this removes only the tail beyond it - zingers,
// overlaps, mis-indexed frames - not good partials. Populated by RefineErrorModel.
bool reject_outliers = false;
double reject_nsigma = 6.0;
std::unordered_map<uint64_t, float> reject_median_I;
size_t reject_count = 0;
bool Mask(const IntegrationOutcome &outcome);
public:
MergeOnTheFly(const DiffractionExperiment &x);
MergeOnTheFly& ReferenceCell(const std::optional<UnitCell> &cell);
MergeOnTheFly& ExcludeIceRings(bool input) { exclude_ice_rings = input; return *this; }
MergeOnTheFly& FilterByImageCC(bool input) { filter_by_image_cc = input; return *this; }
// Fit the global error model from the spread of symmetry-equivalent observations.
// Call once before merging; AddImage then applies it.
void RefineErrorModel(const std::vector<IntegrationOutcome> &outcomes);
[[nodiscard]] bool ErrorModelActive() const { return error_model_active; }
[[nodiscard]] double ErrorModelA() const { return error_model_a; }
[[nodiscard]] double ErrorModelB() const { return error_model_b; }
[[nodiscard]] double ErrorModelChi2() const { return error_model_chi2; }
// Outlier rejection (driven by ScalingSettings::GetOutlierRejectNsigma) reports its count.
[[nodiscard]] size_t RejectedCount() const { return reject_count; }
// image_id is the image's stable identity (its index in the outcomes vector). The CC1/2 half-set
// is a deterministic hash of it, so the split is reproducible run-to-run and independent of the
// order (or threading) of AddImage calls - not a draw from a shared RNG in call order.
void AddImage(const IntegrationOutcome& outcome, int64_t image_id);
// d_min_override, when set, is the effective high-resolution limit for the shell table (used for
// the automatic resolution cutoff computed by the caller); otherwise the manual
// ScalingSettings high-resolution limit stands. The number of shells is ScalingSettings::ReportShellCount.
MergeStatistics MergeStats(const std::vector<MergedReflection> &merged,
const std::vector<IntegrationOutcome> &reflections,
const std::vector<MergedReflection> &reference = {},
std::optional<double> d_min_override = std::nullopt);
std::vector<MergedReflection> ExportReflections();
};
std::vector<MergedReflection> MergeAll(const DiffractionExperiment &x,
const std::vector<IntegrationOutcome> &reflections);
// Pearson CC between one image's corrected intensities (I * image_scale_corr) and a reference set of
// full intensities, over the reflections that would enter the merge (non-ice, within the resolution
// limit, partiality above the floor, finite). {NAN, n} when fewer than 20 reflections qualify.
// This is the per-image image_scale_cc: ScaleOnTheFly sets it, and StillsPartialityRefine recomputes it
// after refining the partiality model, so the reported CC always describes the corrections that will be
// merged - which matters because --min-image-cc drops images by it.
std::pair<double, size_t> ImageReferenceCC(const std::vector<Reflection> &reflections,
const std::map<HKLKey, double> &reference,
const HKLKeyGenerator &generator,
std::optional<double> d_min_limit,
std::optional<double> d_max_limit,
double min_partiality);