PP-C1: drop the v0.4.0 parity tests, keep the histogram goldens (v0.5.1, 1/3)

AGENTS.md

@@ -331,7 +331,7 @@ src/morphobricks_preprocessing/
 notebooks/MorphoBricks_Preprocessing.ipynb   thin notebook: every code cell calls notebook_ui only
 notebooks/notebook_latest_versions.yaml      the notebook's latest version, for LabConstrictor's version check
 tests/              pytest with simulated TIFFs (tests/tiff_tree.py, tests/simulation.py); the
-                    schema-version-2 fixture (tests/v2_fixture.py, tests/data)
+                    histogram science goldens (tests/histogram_goldens.py, tests/data)
 experiments/binning_study/      the study behind the binning decisions (not part of the package)
 ```
 
@@ -534,9 +534,9 @@ functions the tests use.
 - **Rows** are sorted by path, then z, t.
 - **Earlier versions:** the reader refuses schema 1 and 2 (`CsvFormatError`); a run
   recomputes them completely (manifest `schema_version` differs). Each channel's
-  histogram equals the version-2 row's, whole and in blocks
-  (`tests/test_schema_v2_equivalence.py` with the committed fixture in `tests/data`,
-  written by `tests/data/make_v2_fixture.py` with the pre-P3b code).
+  histogram equals the reference row's, whole and in blocks (science goldens:
+  `tests/test_histogram_goldens.py` with the committed reference in `tests/data`, written by
+  `tests/data/make_histogram_goldens.py` with an earlier, independent implementation).
 - **Step 4** summarises each channel of a group on its own (issue #104): one
   `GroupSummary` per group and channel (`channel`, `xyc_plane_count`,
   `channel_histogram_count`), a `channel` column in the table and one plot panel per

tests/data/make_v2_fixture.py → tests/data/make_histogram_goldens.py

@@ -1,4 +1,4 @@
-"""Write the schema-version-2 fixture of ``test_schema_v2_equivalence.py``.
+"""Write the reference histograms (science goldens) of ``test_histogram_goldens.py``.
 
 Run once, from the repository root, with the code of version 2: this repository at
 commit ``55f1074`` (before follow-up P3b) and morphobricks-normalization v0.1.0,
@@ -7,13 +7,13 @@
     git worktree add ../v2 55f1074
     git clone -b v0.1.0 https://github.com/CellMigrationLab/morphobricks-normalization ../n1
     pip install ../n1 tifffile zarr pytest
-    PYTHONPATH=../v2/src python tests/data/make_v2_fixture.py
+    PYTHONPATH=../v2/src python tests/data/make_histogram_goldens.py
 
 It writes, next to this script:
 
-- ``v2_pixels.npz``: the pixels of every image of the experiments in
-  ``tests/v2_fixture.py``;
-- ``v2_<dtype>.csv``: the ``plane_histograms.csv`` that version's
+- ``golden_pixels.npz``: the pixels of every image of the experiments in
+  ``tests/histogram_goldens.py``;
+- ``golden_<dtype>.csv``: the ``plane_histograms.csv`` that version's
   ``run_experiment`` wrote for them (schema version 2, one row per plane and
   channel).
 
@@ -35,8 +35,8 @@
 
 from morphobricks_normalization.csv_io import SCHEMA_VERSION  # noqa: E402
 
+from histogram_goldens import EXPERIMENTS, PIXELS, make_pixels, write_experiment  # noqa: E402
 from morphobricks_preprocessing import prefill_mapping, run_experiment  # noqa: E402
-from v2_fixture import EXPERIMENTS, PIXELS, make_pixels, write_experiment  # noqa: E402
 
 
 def main() -> None:
@@ -52,7 +52,7 @@ def main() -> None:
             result = run_experiment(root, prefill_mapping(axes), bin_count=bins)
             if not result.written:
                 sys.exit(f"{dtype}: {[issue.message for issue in result.issues]}")
-            shutil.copyfile(result.csv_path, HERE / f"v2_{dtype}.csv")
+            shutil.copyfile(result.csv_path, HERE / f"golden_{dtype}.csv")
         print(f"{dtype}: {result.plane_count} rows")
 
 

tests/v2_fixture.py → tests/histogram_goldens.py

@@ -1,8 +1,9 @@
-"""The experiments of the schema-version-2 fixture in ``tests/data`` (follow-up P3b).
+"""The experiments of the reference histograms (science goldens) in ``tests/data``.
 
-``tests/data/make_v2_fixture.py`` wrote their ``plane_histograms.csv`` with the
-code of version 2 (one row per plane and channel); ``test_schema_v2_equivalence.py``
-writes the same images again and compares the current table with it.
+``tests/data/make_histogram_goldens.py`` wrote their ``plane_histograms.csv`` with an earlier,
+independent implementation (one row per plane and channel, before follow-up P3b);
+``test_histogram_goldens.py`` writes the same images again and compares the current table with
+those values.
 """
 
 from __future__ import annotations
@@ -14,7 +15,7 @@
 from tiff_tree import write_data
 
 DATA = Path(__file__).parent / "data"
-PIXELS = DATA / "v2_pixels.npz"
+PIXELS = DATA / "golden_pixels.npz"
 
 # Per pixel type: the bin count, the experiment axis roles and its images as
 # (relative path, axes, writer options). TIFF images share one arrangement of axes;
@@ -74,7 +75,7 @@ def make_pixels(seed: int = 0) -> dict[str, np.ndarray]:
 MorphoBricks v1 analyses 2-D planes (owner decision of 5 November 2026): a Z-stack is
 refused, so the tests write each uint16 image with its planes along T instead. The pixels
 and every histogram are the reference's; only the reference's ``z`` is now ``t``.
-``make_v2_fixture.py`` writes them as the reference was made (``as_made=True``).
+``make_histogram_goldens.py`` writes them as the reference was made (``as_made=True``).
 """