Reorganize tests for readability (#47)

Author: tsalo

test/test_cifti_cli.py

@@ -1,47 +1,13 @@
-"""Unit tests for CIFTI validation helpers."""
+"""Unit tests for CIFTI utility helpers."""
 
 from __future__ import annotations
 
-import nibabel as nb
 import numpy as np
 import pytest
-from nibabel.cifti2.cifti2_axes import BrainModelAxis, ParcelsAxis, ScalarAxis
+from nibabel.cifti2.cifti2_axes import ScalarAxis
+from utils import make_dscalar, make_parcels_axis, make_pconn, make_pscalar
 
-from modelarrayio.utils.cifti import extract_cifti_scalar_data
-
-
-def _make_scalar_cifti(mask_bool: np.ndarray, values: np.ndarray) -> nb.Cifti2Image:
-    scalar_axis = ScalarAxis(['synthetic'])
-    brain_axis = BrainModelAxis.from_mask(mask_bool)
-    header = nb.cifti2.Cifti2Header.from_axes((scalar_axis, brain_axis))
-    return nb.Cifti2Image(values.reshape(1, -1).astype(np.float32), header=header)
-
-
-def _make_parcels_axis(parcel_names: list[str]) -> ParcelsAxis:
-    """Create a minimal surface-only ParcelsAxis for testing."""
-    # One vertex per parcel on the left cortex
-    n = len(parcel_names)
-    nvertices = {'CIFTI_STRUCTURE_CORTEX_LEFT': n}
-    vox_dtype = np.dtype([('ijk', '<i4', (3,))])
-    voxels = [np.array([], dtype=vox_dtype) for _ in range(n)]
-    vertices = [{'CIFTI_STRUCTURE_CORTEX_LEFT': np.array([i], dtype=np.int32)} for i in range(n)]
-    affine = np.eye(4)
-    volume_shape = (10, 10, 10)
-    return ParcelsAxis(parcel_names, voxels, vertices, affine, volume_shape, nvertices)
-
-
-def _make_pscalar_cifti(parcel_names: list[str], values: np.ndarray) -> nb.Cifti2Image:
-    scalar_axis = ScalarAxis(['synthetic'])
-    parcels_axis = _make_parcels_axis(parcel_names)
-    header = nb.cifti2.Cifti2Header.from_axes((scalar_axis, parcels_axis))
-    return nb.Cifti2Image(values.reshape(1, -1).astype(np.float32), header=header)
-
-
-def _make_pconn_cifti(parcel_names: list[str], values: np.ndarray) -> nb.Cifti2Image:
-    parcels_axis = _make_parcels_axis(parcel_names)
-    header = nb.cifti2.Cifti2Header.from_axes((parcels_axis, parcels_axis))
-    n = len(parcel_names)
-    return nb.Cifti2Image(values.reshape(n, n).astype(np.float32), header=header)
+from modelarrayio.utils.cifti import brain_names_to_dataframe, extract_cifti_scalar_data
 
 
 class _FakeHeader:
@@ -66,7 +32,7 @@ def test_extract_cifti_scalar_data_returns_data_and_names() -> None:
     mask = np.zeros((2, 2, 2), dtype=bool)
     mask[0, 0, 0] = True
     mask[1, 1, 1] = True
-    image = _make_scalar_cifti(mask, np.array([1.0, 2.0], dtype=np.float32))
+    image = make_dscalar(mask, np.array([1.0, 2.0], dtype=np.float32))
 
     data, names = extract_cifti_scalar_data(image)
 
@@ -95,7 +61,7 @@ def test_extract_cifti_scalar_data_rejects_inconsistent_reference_names() -> Non
     mask = np.zeros((2, 2, 2), dtype=bool)
     mask[0, 0, 0] = True
     mask[1, 1, 1] = True
-    image = _make_scalar_cifti(mask, np.array([1.0, 2.0], dtype=np.float32))
+    image = make_dscalar(mask, np.array([1.0, 2.0], dtype=np.float32))
 
     with pytest.raises(ValueError, match='Inconsistent greyordinate names'):
         extract_cifti_scalar_data(image, reference_brain_names=np.array(['wrong', 'names']))
@@ -104,7 +70,7 @@ def test_extract_cifti_scalar_data_rejects_inconsistent_reference_names() -> Non
 def test_extract_cifti_scalar_data_pscalar_returns_data_and_names() -> None:
     parcel_names = ['parcel_A', 'parcel_B', 'parcel_C']
     values = np.array([1.0, 2.0, 3.0], dtype=np.float32)
-    image = _make_pscalar_cifti(parcel_names, values)
+    image = make_pscalar(parcel_names, values)
 
     data, names = extract_cifti_scalar_data(image)
 
@@ -115,7 +81,7 @@ def test_extract_cifti_scalar_data_pscalar_returns_data_and_names() -> None:
 def test_extract_cifti_scalar_data_pscalar_validates_reference_names() -> None:
     parcel_names = ['parcel_A', 'parcel_B', 'parcel_C']
     values = np.array([1.0, 2.0, 3.0], dtype=np.float32)
-    image = _make_pscalar_cifti(parcel_names, values)
+    image = make_pscalar(parcel_names, values)
 
     with pytest.raises(ValueError, match='Inconsistent parcel names'):
         extract_cifti_scalar_data(image, reference_brain_names=np.array(['X', 'Y', 'Z']))
@@ -125,7 +91,7 @@ def test_extract_cifti_scalar_data_pconn_flattens_matrix() -> None:
     parcel_names = ['parcel_A', 'parcel_B']
     n = len(parcel_names)
     matrix = np.arange(n * n, dtype=np.float32).reshape(n, n)
-    image = _make_pconn_cifti(parcel_names, matrix)
+    image = make_pconn(parcel_names, matrix)
 
     data, names = extract_cifti_scalar_data(image)
 
@@ -140,7 +106,7 @@ def test_extract_cifti_scalar_data_pconn_validates_reference_names() -> None:
     parcel_names = ['parcel_A', 'parcel_B']
     n = len(parcel_names)
     matrix = np.zeros((n, n), dtype=np.float32)
-    image = _make_pconn_cifti(parcel_names, matrix)
+    image = make_pconn(parcel_names, matrix)
 
     # Get the correct element_names first
     _, element_names = extract_cifti_scalar_data(image)
@@ -150,3 +116,23 @@ def test_extract_cifti_scalar_data_pconn_validates_reference_names() -> None:
     bad_names[0] = 'wrong'
     with pytest.raises(ValueError, match='Inconsistent parcel names'):
         extract_cifti_scalar_data(image, reference_brain_names=bad_names)
+
+
+def test_make_parcels_axis_produces_valid_axis() -> None:
+    """Smoke test: make_parcels_axis should return a ParcelsAxis with correct length."""
+    from nibabel.cifti2.cifti2_axes import ParcelsAxis
+
+    names = ['A', 'B', 'C']
+    axis = make_parcels_axis(names)
+    assert isinstance(axis, ParcelsAxis)
+    assert len(axis) == len(names)
+
+
+def test_brain_names_to_dataframe() -> None:
+    names = np.array(['CORTEX_LEFT', 'CORTEX_LEFT', 'CORTEX_RIGHT'])
+    gdf, struct_strings = brain_names_to_dataframe(names)
+    assert len(gdf) == 3
+    assert 'vertex_id' in gdf.columns
+    assert 'structure_id' in gdf.columns
+    assert gdf['vertex_id'].tolist() == [0, 1, 2]
+    assert len(struct_strings) == 2  # factorize unique structures

test/test_cifti_cli_comprehensive.py

@@ -1,3 +1,7 @@
+"""Tests for the mif-to-h5 CLI command."""
+
+from __future__ import annotations
+
 import csv
 
 import h5py

test/test_cifti_utils.py

@@ -1,4 +1,4 @@
-"""Unit tests for fixel utility error handling."""
+"""Unit tests for MIF/fixel utility helpers."""
 
 from __future__ import annotations
 

test/test_mif_cli.py

@@ -1,12 +1,10 @@
-"""Tests for CIFTI cohort normalization and greyordinate helpers."""
+"""Unit tests for modelarrayio.utils.misc cohort and scalar-source helpers."""
 
 from __future__ import annotations
 
-import numpy as np
 import pandas as pd
 import pytest
 
-from modelarrayio.utils.cifti import brain_names_to_dataframe
 from modelarrayio.utils.misc import build_scalar_sources, cohort_to_long_dataframe
 
 
@@ -74,13 +72,3 @@ def test_build_scalar_sources_ordering() -> None:
     assert list(src.keys()) == ['A', 'B']
     assert src['A'] == ['x1', 'x2']
     assert src['B'] == ['y1']
-
-
-def test_brain_names_to_dataframe() -> None:
-    names = np.array(['CORTEX_LEFT', 'CORTEX_LEFT', 'CORTEX_RIGHT'])
-    gdf, struct_strings = brain_names_to_dataframe(names)
-    assert len(gdf) == 3
-    assert 'vertex_id' in gdf.columns
-    assert 'structure_id' in gdf.columns
-    assert gdf['vertex_id'].tolist() == [0, 1, 2]
-    assert len(struct_strings) == 2  # factorize unique structures

test/test_fixels_utils.py test/test_mif_utils.pytest/test_fixels_utils.py renamed to test/test_mif_utils.py

@@ -1,26 +1,26 @@
+"""Tests for the nifti-to-h5 and h5-to-nifti CLI commands."""
+
+from __future__ import annotations
+
 import csv
+import logging
 import os.path as op
 
 import h5py
 import nibabel as nb
 import numpy as np
 import pytest
 import tiledb
+from utils import make_nifti
 
 from modelarrayio.cli.main import main as modelarrayio_main
 
 
-def _make_nifti(data, affine=None):
-    if affine is None:
-        affine = np.eye(4)
-    return nb.Nifti1Image(data.astype(np.float32), affine)
-
-
 def _ijk_value(i, j, k):
     return i * 100.0 + j * 10.0 + k * 1.0
 
 
-def test_convoxel_cli_creates_expected_hdf5(tmp_path, monkeypatch):
+def test_nifti_to_h5_creates_expected_hdf5(tmp_path, monkeypatch):
     # Small synthetic volume
     shape = (5, 6, 7)
     group_mask = np.zeros(shape, dtype=bool)
@@ -30,7 +30,7 @@ def test_convoxel_cli_creates_expected_hdf5(tmp_path, monkeypatch):
         group_mask[i, j, k] = True
 
     # Save group mask
-    group_mask_img = _make_nifti(group_mask.astype(np.uint8))
+    group_mask_img = make_nifti(group_mask.astype(np.uint8))
     group_mask_file = tmp_path / 'group_mask.nii.gz'
     group_mask_img.to_filename(group_mask_file)
 
@@ -48,8 +48,8 @@ def test_convoxel_cli_creates_expected_hdf5(tmp_path, monkeypatch):
             omit = true_coords[1]
             indiv_mask[omit] = False
 
-        scalar_img = _make_nifti(scalar)
-        mask_img = _make_nifti(indiv_mask.astype(np.uint8))
+        scalar_img = make_nifti(scalar)
+        mask_img = make_nifti(indiv_mask.astype(np.uint8))
 
         scalar_path = tmp_path / f'sub-{sidx + 1}_scalar.nii.gz'
         mask_path = tmp_path / f'sub-{sidx + 1}_mask.nii.gz'
@@ -144,15 +144,15 @@ def test_convoxel_cli_creates_expected_hdf5(tmp_path, monkeypatch):
             assert np.isclose(v1, expected_s1, equal_nan=True)
 
 
-def test_h5_to_nifti_cli_writes_results_with_dataset_column_names(tmp_path):
+def test_h5_to_nifti_writes_results_with_dataset_column_names(tmp_path):
     shape = (3, 3, 3)
     group_mask = np.zeros(shape, dtype=bool)
     true_coords = [(0, 0, 0), (1, 1, 1), (2, 2, 2)]
     for coord in true_coords:
         group_mask[coord] = True
 
     group_mask_file = tmp_path / 'group_mask.nii.gz'
-    _make_nifti(group_mask.astype(np.uint8)).to_filename(group_mask_file)
+    make_nifti(group_mask.astype(np.uint8)).to_filename(group_mask_file)
 
     in_file = tmp_path / 'results.h5'
     with h5py.File(in_file, 'w') as h5:
@@ -215,12 +215,12 @@ def test_nifti_to_h5_scalar_columns_writes_prefixed_outputs(tmp_path, monkeypatc
         group_mask[i, j, k] = True
 
     group_mask_file = tmp_path / 'group_mask.nii.gz'
-    _make_nifti(group_mask.astype(np.uint8)).to_filename(group_mask_file)
+    make_nifti(group_mask.astype(np.uint8)).to_filename(group_mask_file)
 
     rows = []
     for sidx in range(2):
         subj_mask_file = tmp_path / f'sub-{sidx + 1}_mask.nii.gz'
-        _make_nifti(group_mask.astype(np.uint8)).to_filename(subj_mask_file)
+        make_nifti(group_mask.astype(np.uint8)).to_filename(subj_mask_file)
 
         alpha_data = np.zeros(shape, dtype=np.float32)
         beta_data = np.zeros(shape, dtype=np.float32)
@@ -230,8 +230,8 @@ def test_nifti_to_h5_scalar_columns_writes_prefixed_outputs(tmp_path, monkeypatc
 
         alpha_file = tmp_path / f'sub-{sidx + 1}_alpha.nii.gz'
         beta_file = tmp_path / f'sub-{sidx + 1}_beta.nii.gz'
-        _make_nifti(alpha_data).to_filename(alpha_file)
-        _make_nifti(beta_data).to_filename(beta_file)
+        make_nifti(alpha_data).to_filename(alpha_file)
+        make_nifti(beta_data).to_filename(beta_file)
 
         rows.append(
             {
@@ -293,7 +293,7 @@ def _build_nifti_cohort(tmp_path):
     for coord in true_coords:
         group_mask[coord] = 1
     group_mask_file = tmp_path / 'group_mask.nii.gz'
-    _make_nifti(group_mask).to_filename(group_mask_file)
+    make_nifti(group_mask).to_filename(group_mask_file)
 
     rows = []
     for sidx in range(2):
@@ -302,8 +302,8 @@ def _build_nifti_cohort(tmp_path):
             scalar[i, j, k] = float(i + j + k + sidx)
         scalar_file = tmp_path / f'sub-{sidx + 1}_scalar.nii.gz'
         mask_file = tmp_path / f'sub-{sidx + 1}_mask.nii.gz'
-        _make_nifti(scalar).to_filename(scalar_file)
-        _make_nifti(group_mask).to_filename(mask_file)
+        make_nifti(scalar).to_filename(scalar_file)
+        make_nifti(group_mask).to_filename(mask_file)
         rows.append(
             {
                 'scalar_name': 'FA',
@@ -321,15 +321,13 @@ def _build_nifti_cohort(tmp_path):
     return group_mask_file, cohort_csv
 
 
-def test_nifti_tiledb_fails_when_output_already_exists(tmp_path, monkeypatch, caplog):
+def test_nifti_tiledb_removes_existing_arrays_on_rerun(tmp_path, monkeypatch, caplog):
     """Regression test for https://github.com/PennLINC/ModelArrayIO/issues/39.
 
     The TileDB backend should succeed when the output directory already contains
     arrays from a previous run, removing and recreating them, and should emit a
     warning for each removed array.
     """
-    import logging
-
     group_mask_file, cohort_csv = _build_nifti_cohort(tmp_path)
     out_tdb = tmp_path / 'out.tdb'
     monkeypatch.chdir(tmp_path)

test/test_cifti_cohort.py test/test_misc_utils.pytest/test_cifti_cohort.py renamed to test/test_misc_utils.py

@@ -9,10 +9,6 @@
 from modelarrayio.utils.nifti import flattened_image
 
 
-def _eye_affine():
-    return np.eye(4)
-
-
 def test_flattened_image_extracts_group_masked_values() -> None:
     shape = (3, 3, 3)
     group_mask = np.zeros(shape, dtype=bool)
@@ -24,8 +20,8 @@ def test_flattened_image_extracts_group_masked_values() -> None:
     scalar[2, 0, 1] = 2.5
 
     indiv_mask = group_mask.copy()
-    scalar_img = nb.Nifti1Image(scalar, _eye_affine())
-    mask_img = nb.Nifti1Image(indiv_mask.astype(np.float32), _eye_affine())
+    scalar_img = nb.Nifti1Image(scalar, np.eye(4))
+    mask_img = nb.Nifti1Image(indiv_mask.astype(np.float32), np.eye(4))
 
     flat = flattened_image(scalar_img, mask_img, group_mask)
     assert flat.shape == (2,)
@@ -43,8 +39,8 @@ def test_flattened_image_nan_outside_individual_mask() -> None:
     indiv = group_mask.copy()
     indiv[1, 1, 1] = False
 
-    scalar_img = nb.Nifti1Image(scalar, _eye_affine())
-    mask_img = nb.Nifti1Image(indiv.astype(np.float32), _eye_affine())
+    scalar_img = nb.Nifti1Image(scalar, np.eye(4))
+    mask_img = nb.Nifti1Image(indiv.astype(np.float32), np.eye(4))
 
     flat = flattened_image(scalar_img, mask_img, group_mask)
     assert flat.shape == (2,)

test/test_voxels_cli.py

@@ -0,0 +1,48 @@
+"""Shared test utility functions for ModelArrayIO tests."""
+
+from __future__ import annotations
+
+import nibabel as nb
+import numpy as np
+from nibabel.cifti2.cifti2_axes import BrainModelAxis, ParcelsAxis, ScalarAxis
+
+
+def make_parcels_axis(parcel_names: list[str]) -> ParcelsAxis:
+    """Create a minimal surface-only ParcelsAxis (one vertex per parcel on the left cortex)."""
+    n = len(parcel_names)
+    nvertices = {'CIFTI_STRUCTURE_CORTEX_LEFT': n}
+    vox_dtype = np.dtype([('ijk', '<i4', (3,))])
+    voxels = [np.array([], dtype=vox_dtype) for _ in range(n)]
+    vertices = [{'CIFTI_STRUCTURE_CORTEX_LEFT': np.array([i], dtype=np.int32)} for i in range(n)]
+    return ParcelsAxis(parcel_names, voxels, vertices, np.eye(4), (10, 10, 10), nvertices)
+
+
+def make_dscalar(mask_bool: np.ndarray, values: np.ndarray) -> nb.Cifti2Image:
+    """Create a synthetic dscalar CIFTI image from a volumetric boolean mask and values."""
+    scalar_axis = ScalarAxis(['synthetic'])
+    brain_axis = BrainModelAxis.from_mask(mask_bool)
+    header = nb.cifti2.Cifti2Header.from_axes((scalar_axis, brain_axis))
+    return nb.Cifti2Image(values.reshape(1, -1).astype(np.float32), header=header)
+
+
+def make_pscalar(parcel_names: list[str], values: np.ndarray) -> nb.Cifti2Image:
+    """Create a synthetic pscalar CIFTI image with the given parcel names and values."""
+    scalar_axis = ScalarAxis(['synthetic'])
+    parcels_axis = make_parcels_axis(parcel_names)
+    header = nb.cifti2.Cifti2Header.from_axes((scalar_axis, parcels_axis))
+    return nb.Cifti2Image(values.reshape(1, -1).astype(np.float32), header=header)
+
+
+def make_pconn(parcel_names: list[str], values: np.ndarray) -> nb.Cifti2Image:
+    """Create a synthetic pconn CIFTI image; values are reshaped to (n_parcels, n_parcels)."""
+    parcels_axis = make_parcels_axis(parcel_names)
+    header = nb.cifti2.Cifti2Header.from_axes((parcels_axis, parcels_axis))
+    n = len(parcel_names)
+    return nb.Cifti2Image(values.reshape(n, n).astype(np.float32), header=header)
+
+
+def make_nifti(data: np.ndarray, affine: np.ndarray | None = None) -> nb.Nifti1Image:
+    """Create a Nifti1Image with the given data and an optional affine (default: identity)."""
+    if affine is None:
+        affine = np.eye(4)
+    return nb.Nifti1Image(data.astype(np.float32), affine)