Skip to content

Rescale and LOT correction#424

Open
Soorya19Pradeep wants to merge 34 commits into
mainfrom
rescale-lot-correction
Open

Rescale and LOT correction#424
Soorya19Pradeep wants to merge 34 commits into
mainfrom
rescale-lot-correction

Conversation

@Soorya19Pradeep

@Soorya19Pradeep Soorya19Pradeep commented May 19, 2026

Copy link
Copy Markdown
Contributor

The following have been implemented in this branch:

  • TripletDataModule that rescales the patches to a reference pixel and a uses the Z-MIP projection transform
  • computation and use of the computed linear optical transfer function for batch correction between datasets.
  • a simple case of computation of Maximum Mean Discrepancy (MMD) with sample config.

@edyoshikun
edyoshikun force-pushed the rescale-lot-correction branch from a93d071 to 37b81ff Compare June 29, 2026 17:59
@edyoshikun
edyoshikun changed the base branch from modular-viscy-staging to main June 29, 2026 20:22
Soorya19Pradeep and others added 6 commits June 29, 2026 13:28
…staging

Scale-aware patch extraction (packages/viscy-data):
- Add _read_pixel_size() helper to read X pixel size from OME-Zarr metadata
- Add reference_pixel_size parameter to TripletDataModule: when set, computes
  initial_yx_patch_size from the pixel-size ratio so the same physical area is
  covered at inference time
- Replace BatchedRescaleYXd (removed per review) with existing BatchedZoomd
  using scale_factor=(1.0, final_y/initial_y, final_x/initial_x) and
  mode="bilinear" with antialias; import is lazy to avoid a hard dep

LOT batch correction (applications/dynaclr):
- Add dynaclr.evaluation.lot_correction submodule with core logic (fit, apply,
  save, load), Pydantic configs, and Click CLI entry points
- Register fit-lot-correction and apply-lot-correction in cli.py via LazyCommand
- Add pot and joblib to optional-dependencies.eval in pyproject.toml

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Plots n cell patches side-by-side:
- Left: raw patch at initial_yx_patch_size (larger physical area)
- Right: same patch bilinearly downscaled to final_yx_patch_size (model input)

Both columns share the same percentile contrast window so differences are
spatial. Physical scale (µm × µm) is shown in each panel title.

Usage:
  python visualize_triplet_rescaling.py       --data-path /path/to/data.zarr       --tracks-path /path/to/tracks       --source-channel Phase3D       --z-range 0 5       --final-yx-patch-size 224 224       --reference-pixel-size 0.325       --output rescaling_comparison.png

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…meta

_collate_norm_meta iterated every normalization level and called
torch.stack on each stat, but timepoint_statistics is nested
{timepoint: {stat: tensor}} rather than flat {stat: tensor}. Any zarr
carrying timepoint_statistics (alongside fov/dataset stats) crashed
batch collation in TripletDataModule with "expected Tensor as element 0
... but got dict". Stack within each timepoint sub-dict instead.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…etDataModule

Lets a 3D OME-Zarr feed a 2D model without materializing a separate MIP
dataset, and centers the extracted Z window on each FOV's focus plane.

- z_reduction ("mip"/"center"): collapse the extracted z_range to one
  slice via BatchedChannelWiseZReductiond. Label-free channels (resolved
  by parse_channel_name) take the center slice; others are max-projected.
  on_after_batch_transfer expects Z=1 when reduction is on.
- z_extraction_window/z_focus_offset/focus_channel: resolve a per-FOV
  focus-centered window from each position's
  focus_slice[ch].fov_statistics.z_focus_mean (fallback z_total//2), all
  windows the same width. z_range stays as an explicit override; exactly
  one of z_range / z_extraction_window must be given. Per-FOV windows are
  resolved at setup() and looked up per patch in the dataset.

Tests cover both reduction strategies (discriminating center vs MIP),
the normalize-then-reduce order, per-FOV focus resolution, and the
z_range/z_extraction_window XOR guard.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Documents the TripletDataModule predict path (zarr + tracking, not
parquet) and adds a runnable sample config demonstrating z_reduction +
reference_pixel_size to feed a 3D dataset to a 2D model.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Notebook-style script (no CLI) that loads a 3D OME-Zarr + tracking,
extracts a per-FOV focus-centered Z window, collapses it via z_reduction,
applies a random affine so anchor/positive diverge, and visualizes a
couple of batches to a PNG.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

This PR expands the Triplet-based data loading path to better support cross-dataset inference (pixel-size aware patch extraction + on-the-fly Z reduction + focus-centered per-FOV Z windows), and adds a LOT (Linear Optimal Transport) batch-correction pipeline with CLI entrypoints for embedding-space correction.

Changes:

  • Extend TripletDataModule/TripletDataset to support per-FOV z_range resolution via z_extraction_window + focus_slice metadata, plus optional z_reduction and pixel-size aware rescaling.
  • Add LOT correction core functions + Pydantic config models + dynaclr CLI commands for fitting/applying correction pipelines.
  • Add tests for the new Triplet Z-window/Z-reduction behavior and update docs/config examples for inference.

Reviewed changes

Copilot reviewed 14 out of 14 changed files in this pull request and generated 7 comments.

Show a summary per file
File Description
packages/viscy-data/src/viscy_data/triplet.py Adds pixel-size rescaling, per-FOV focus-centered Z windows, and optional Z-reduction in the Triplet datamodule/dataset.
packages/viscy-data/tests/test_triplet.py Adds tests for z_extraction_window XOR validation, per-FOV focus-centered windows, and Z-reduction behavior/order.
packages/viscy-data/src/viscy_data/_utils.py Extends norm-meta collation to correctly stack nested timepoint_statistics.
applications/dynaclr/src/dynaclr/evaluation/lot_correction/lot_correction.py Implements LOT correction fit/apply + pipeline save/load helpers.
applications/dynaclr/src/dynaclr/evaluation/lot_correction/config.py Adds Pydantic models for fit/apply configs and filter specs.
applications/dynaclr/src/dynaclr/evaluation/lot_correction/fit_lot_correction.py CLI wrapper for fitting and saving a LOT pipeline from YAML config.
applications/dynaclr/src/dynaclr/evaluation/lot_correction/apply_lot_correction.py CLI wrapper for applying a saved LOT pipeline to an embedding zarr.
applications/dynaclr/src/dynaclr/evaluation/lot_correction/init.py Exposes LOT correction functions at the package level.
applications/dynaclr/src/dynaclr/cli.py Registers new fit-lot-correction and apply-lot-correction CLI subcommands.
applications/dynaclr/pyproject.toml Adds joblib + pot to eval extras for LOT correction dependencies.
applications/dynaclr/scripts/dataloader_inspection/visualize_triplet_rescaling.py Adds a visualization utility for the new pixel-size rescaling behavior.
applications/dynaclr/scripts/dataloader_inspection/triplet_dataloader_zprojection.py Adds an inspection notebook/script for focus-centered Z windows + Z-reduction.
applications/dynaclr/docs/DAGs/inference_triplet.md Documents Triplet inference path and new options (needs a small wording update for z-window semantics).
applications/dynaclr/configs/prediction/predict_triplet_2d_from_3d.yml Provides a sample prediction config demonstrating reference_pixel_size + z_reduction.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Comment thread packages/viscy-data/tests/test_triplet.py
Comment thread packages/viscy-data/src/viscy_data/triplet.py Outdated
Comment thread packages/viscy-data/src/viscy_data/triplet.py
Comment thread packages/viscy-data/src/viscy_data/triplet.py
Comment thread packages/viscy-data/src/viscy_data/triplet.py
Comment thread applications/dynaclr/docs/DAGs/inference_triplet.md Outdated
edyoshikun and others added 17 commits June 29, 2026 14:07
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
- Declare viscy-transforms in viscy-data's triplet extra and hoist the
  BatchedZoomd/BatchedChannelWiseZReductiond imports to module top; the
  inline imports were masking a missing runtime dependency that would
  ImportError for anyone using z_reduction or reference_pixel_size.
- Raise in _resolve_per_fov_z_ranges when a FOV's Z is smaller than
  z_extraction_window, instead of silently emitting a narrower window
  that breaks cross-FOV batch stacking.
- Remove a dead duplicate break in test_focus_centered_z_range.
- Correct the inference DAG doc: embeddings live in .X (the embedding_key
  array), mirrored to obsm["X_backbone"]/["X_projections"].

uv.lock intentionally omitted: the workspace glob currently entangles the
untracked applications/eet package, so a clean regen of the single
viscy-transforms edge is not possible until eet is committed or removed.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Replace the NumPy/scipy.cdist backend with PyTorch so the pooled RBF
kernel matrix is built once on the available device (CUDA if usable,
else CPU) and reused across all permutations. Device selection probes a
trivial kernel launch and falls back to CPU, so a present-but-unusable
GPU (e.g. compute capability older than the torch build) does not crash.

Public API is unchanged: median_heuristic/compute_mmd_unbiased return the
same bandwidth-convention values, and mmd_permutation_test still returns
(mmd2, p_value, null_distribution) so dynaclr's effect-size and activity
z-score computations keep working.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Add witness_function to the MMD module: the per-point empirical witness
w(z) = mean_i k(z, x_i) - mean_j k(z, y_j), the RKHS direction along
which distributions P (X) and Q (Y) differ. Positive scores lean toward
X, negative toward Y. Reuses the GPU RBF-kernel helpers and bandwidth
(median-heuristic) convention already in the module, and chunks over
query points to bound the intermediate kernel matrices.

Tests cover distribution separation, X<->Y antisymmetry, and chunk-size
invariance.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adds `label_source: witness` to the linear-classifier eval as a second
option to annotation-CSV labels. Per marker, cells are pooled across
experiments, control/perturbed references are built from per-experiment
wells, the MMD witness is fit and scored per cell, and scores are gated
(sign + dead-zone) into control/perturbed pseudo-labels. Everything
downstream (train_linear_classifier, publish, append-predictions, plots)
is unchanged — only the label source differs.

- evaluate_config.py: WitnessLabelSource, WitnessSettings, label_source
  switch on LinearClassifiersStepConfig (annotations stays the default).
- witness_labels.py: well->reference mask (path-component match so C/1
  != C/10), witness fit/score, dead-zone gating.
- orchestrated.py: branch label assembly into _annotation_run_specs /
  _witness_run_specs / _build_labeled_adata; training loop unchanged.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Five integration tests exercising the real code path:
- well-prefix match rejects C/1 vs C/10 confusion
- witness scores separate control/perturbed clusters
- dead-zone drops ambiguous near-origin cells
- empty AnnData when a reference group has no cells
- end-to-end run_linear_classifiers witness mode → metrics + summary PDF

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Mirrors linear_classifiers_infectomics.yml but with label_source: witness.
Per-experiment control/perturbed wells replace annotation CSVs; wells are
templates to edit per plate layout. One classifier per marker (G3BP1,
SEC61B, Phase3D, viral_sensor), dead_zone 0.1, track-level split.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- witness_score_classifiers.md: full DAG (why, step-by-step, gating rule,
  config, notes) plus embedded visuals.
- visuals/: three Graphviz flow diagrams (data flow, gating, pipeline) as
  .dot + png + pdf, and four mock example-output plots (witness-score
  histogram, per-marker metrics bar, ROC, F1-over-time) with a reproducible
  generator (mock_witness_plots.py). Mock plots are watermarked synthetic.
- Cross-links from evaluation.md and linear_classifiers/README.md so the
  new DAG is discoverable from the annotation path.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The witness label is a deterministic function of the embedding, so
evaluating a witness-trained classifier against that same label is
circular — it reports a trivial ~1.0 (verified on real infectomics
embeddings: 1.000 vs witness label, 0.712 vs true infection_state).

WitnessSettings gains eval_against (default "infection_state") and
eval_class_map ({control: uninfected, perturbed: infected}). In witness
mode, val metrics are recomputed on the val split against the ground-truth
obs column via the class map; the summary records eval_source. When the
column is absent, metrics fall back to the witness label flagged as
eval_source="witness_label". Adds two tests (annotation eval + fallback).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Score the witness-trained classifier against infection_state (not the
self-referential witness label) with the control→uninfected /
perturbed→infected class map, matching the leakage fix.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- witness_score_classifiers.md: add "Evaluation: avoid the circularity
  trap" section with the real per-marker metrics table (scored vs
  infection_state), eval_against/eval_class_map in the config block, and
  step 7 in the flow; refresh the Notes caveat (metrics are honest now,
  check eval_source; note class imbalance).
- witness_dataflow.dot/png/pdf: add the EVALUATE-vs-annotations node.
- mock_witness_plots.py + bar/ROC png/pdf: use real run values (viral_sensor
  0.87, SEC61B 0.76, Phase3D 0.58, G3BP1 0.49) instead of invented all-high
  numbers; watermark clarifies bar/ROC are real, hist/F1 synthetic.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
edyoshikun and others added 11 commits July 14, 2026 20:15
Witness mode produced only the standard metrics/ROC/F1 plots — nothing
showing HOW the pseudo-labels were chosen. build_witness_labels now
attaches the gating diagnostic to the returned AnnData (obs["witness_score"],
obs["witness_ref"] = control_well/perturbed_well/other, uns["witness_gating"]
with threshold, bandwidth, counts, and the pre-gating score/ref arrays).

_save_task_plots leads the {task}_summary.pdf with a per-marker gating page:
the witness-score histogram split by well-of-origin, the dropped dead-zone
band, the sign cut, and labeled/dropped counts. Verified on real SEC61B
embeddings — control/perturbed wells separate weakly, most labeled cells are
"other" near zero, which explains the modest AUROC at a glance.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Two related extensions to the witness LC label source:

1. Filter-based references. WitnessLabelSource gains control_filter /
   perturbed_filter (arbitrary obs filters) as an alternative to
   control_wells / perturbed_wells. obs_filter_mask supports scalar (==),
   list (isin), and range dicts ({lt,le,gt,ge}, one bound = half-line, two =
   window), with a well/fov_name key routing to the well-prefix match. This
   enables contrasts like early-vs-late timepoints or control-vs-perturbed-at-
   late as weak-label sources. Validator: wells XOR filter per side, no mixing.

2. Annotation-scored eval. WitnessSettings.eval_annotations lets a witness run
   (weak labels for training) be scored against real infection_state joined
   from annotation CSVs when the embeddings obs lacks the column, instead of
   the perturbation proxy. _join_eval_annotations does the per-experiment join.

Tests: obs_filter_mask forms + filter-based late-window integration (8 green).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…tness label

The leakage fix recomputed the metrics CSV against ground-truth infection_state
but left the ROC + F1-over-time pages drawing from the trainer's witness-label
val arrays — so the summary PDF showed a circular AUROC=1.000 while the metrics
bar/CSV correctly showed ~0.80. _evaluate_witness_against_annotations now also
returns the annotation-scored y_val / y_val_proba / classes / val_hours (aligned
to the has-ground-truth subset), and the orchestrator swaps them into the
plotting outputs so the whole PDF is consistent.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The witness axis measures how much a MARKER's embedding changes between the
references, so its meaning is marker-dependent: viral_sensor -> infection,
organelle markers (SEC61B/TOMM20/G3BP1) -> remodeling. Scoring every marker
against infection_state was wrong for the organelle markers.

WitnessSettings.marker_eval maps a marker to {eval_against, eval_class_map},
overriding the top-level target for that marker (absent markers fall back).
_resolve_witness_eval applies the override per run; eval_source in the summary
records the actual target used. Lets one run score viral_sensor vs
infection_state and SEC61B vs organelle_state.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Replace the sign+dead-zone witness gate and its circular annotation grading
with a two-stage, annotation-format pipeline (Soorya Pradeep's recipe,
framework-ized). A witness->GMM label IS an annotation: its meaning is named by
the modality it is computed from (viral_sensor -> infection_state; organelle ->
organelle_remodeling_state), so it flows through the existing annotation training
path unchanged.

Stage A (new `dynaclr witness-gmm-labels`): witness score -> per-condition
2-component GMM -> writes an annotation file (named state column + real class
vocabulary). Negatives = all control-well cells; positives = perturbed cells
with GMM posterior >= threshold; unimodal (near-noise) markers skipped.

Stage B: the existing `run-linear-classifiers` (label_source: annotations),
untouched. Teacher/student is free — point Stage B at a different modality's
zarr (e.g. SEC61 labels -> train on phase), joined by cell key.

- New shared `viscy_utils.evaluation.witness_gmm.fit_gmm_labels` (+ tests).
- Rename witness_labels{,_test}.py -> witness_gmm_labels{,_test}.py; recipe ->
  witness_gmm_labels_infectomics.yml.
- Remove WitnessSettings, _gate_scores, _evaluate_witness_against_annotations,
  _resolve_witness_eval, _join_eval_annotations, the label_source: witness
  branch, and the superseded witness_score_classifiers DAG + visuals.
- New DAG witness_gmm_classifiers.md; update evaluation.md + LC README.
- Validated: fit_gmm_labels reproduces Soorya's saved GMM on
  2026_03_24_A549_SEC61_ZIKV (bandwidth 238~240, remod weight 0.655~0.649,
  confident-pos 59.1%~58.4%).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Remove generated mock visualization PNG/PDF files from version control.
Files remain on disk but are no longer tracked.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Collapse _get_device() (~20 lines + module cache) to a one-line _DEVICE
  constant; torch.cuda.is_available() is sufficient here.
- Remove the unused public gaussian_rbf_kernel() NumPy wrapper; nothing
  imports it and the internal _rbf_kernel does all the work.
- Hoist the duplicated _subsample() (identical in compute_mmd.py and
  witness_gmm_labels.py) into a single shared viscy_utils.evaluation.mmd.subsample.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
_read_pixel_size now raises on a zero/negative/NaN OME-Zarr X scale
instead of letting reference/inference division produce inf/NaN that
silently corrupts initial_yx_patch_size.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
_collate_norm_meta took timepoint keys from the first sample and indexed
every other sample with them; a batch mixing FOVs whose zattrs expose
different timepoint_statistics sets raised a cryptic KeyError. Validate
the key sets up front and raise an actionable message instead.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The classic TripletDataModule path returned the full per-FOV norm_meta
from zattrs without selecting the sample's timepoint, so
NormalizeSampled(level='timepoint_statistics') received the nested
{tp_idx: {stat: Tensor}} layout and raised KeyError('mean'). Live
DynaCLR configs (DynaCLR-3D-BagOfChannels-v2, DynaCLR-2D-MIP-BagOfChannels,
DINOv3-temporal) use this level, but only via MultiExperimentDataModule,
which already pre-resolves — so the classic path was latently broken.

- Hoist the timepoint resolver out of SlidingWindowDataset into a shared
  _utils._resolve_timepoint_norm_meta and reuse it in both datasets.
- Resolve the timepoint in TripletDataset._slice_patch using the row's t.
- Drop the now-dead nested timepoint_statistics branch in
  _collate_norm_meta (all collated metas are flat after resolution). This
  supersedes the mixed-timepoint guard added in 9b0a114.
- Add a triplet integration test that fails without the resolution and
  point the existing sliding-window helper test at the shared function.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
viscy_data eagerly imports triplet.py at package root, so a module-level
`from viscy_transforms import ...` made viscy-transforms a hard
requirement of `import viscy_data` — forcing the working tree to add it
to core dependencies even for HCS-only users.

Move the two imports (BatchedZoomd, BatchedChannelWiseZReductiond) into
the reference_pixel_size / z_reduction branches where they are used, so
they load only when those features are enabled. viscy-transforms stays in
the [triplet] extra; `import viscy_data` no longer requires it.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants