pycroglia.core.tests.unit package

Submodules

pycroglia.core.tests.unit.test_arclength.test_circle_linear()

Test arc length of a circle using linear chords.

This test discretizes a unit circle into 20 points and computes its arc length using straight-line (linear chordal) distances. The result is expected to underestimate the true circumference slightly, but remain within 2% relative tolerance.

Asserts:

The computed arc length is close to 2π within rtol=0.02.

pycroglia.core.tests.unit.test_arclength.test_circle_pchip()

Test arc length of a circle using PCHIP interpolation.

This test discretizes a unit circle into 20 points and computes its arc length using Piecewise Cubic Hermite Interpolating Polynomial (PCHIP) interpolation. The smoother parametric curve gives a more accurate estimate of the circumference.

Asserts:

The computed arc length is close to 2π within rtol=0.005.

pycroglia.core.tests.unit.test_arclength.test_circle_spline()

Test arc length of a circle using cubic spline interpolation.

This test discretizes a unit circle into 20 points and computes its arc length using cubic spline interpolation. The smooth spline fit should provide a highly accurate estimate of the circumference.

Asserts:

The computed arc length is close to 2π within rtol=0.005.

pycroglia.core.tests.unit.test_bounding_box.test_bounding_box_of_cell()

Test Bounding Box computation on a simple 3D volume.

Scenario:

A single foreground voxel is placed at (z=2, y=3, x=4) in a (5,6,7) volume. The bounding box should crop along Y and X to tightly include only that voxel.

Asserts:

• The cropped volume has shape (5, 1, 1), shrinking only in y and x.

• The bounding box left/right (x bounds) are both 4.

• The bounding box bottom/top (y bounds) are both 3.

• The voxel at the expected cropped coordinate is True.

pycroglia.core.tests.unit.test_branch_analysis.indices_to_mask(cell_indices, img_shape)

Convert a 1D list of voxel indices into a 3D binary mask.

Parameters:

• cell_indices (np.ndarray) – 1D array of flattened (raveled) voxel indices where the mask should be True.

• img_shape (tuple[int, int, int]) – Shape of the desired output mask, given as (Z, Y, X).

Returns:

3D uint8 mask of shape img_shape with True (1) at all specified voxel indices and False (0) elsewhere.

Return type:

np.ndarray

pycroglia.core.tests.unit.test_branch_analysis.test_branch_analysis_equivalence()

Test BranchAnalysis matches MATLAB reference results.

Asserts:

• Python and MATLAB endpoint voxel masks are identical.

• The number of detected branch points matches MATLAB output.

• Branch point coordinates match element-wise.

• Maximum, minimum, and average branch lengths match within 1e-6.

pycroglia.core.tests.unit.test_centroid.test_compute_average_centroid_distance()

Test compute_average_distance returns correct scaled distance.

Asserts:

The average pairwise distance between two centroids is correctly computed in physical units, given voxel scales for XY and Z.

pycroglia.core.tests.unit.test_centroid.test_get_centroids()

Test Centroids correctly computes centroids from binary masks.

Asserts:

The computed centroids match the expected voxel coordinates for two simple 3D masks.

pycroglia.core.tests.unit.test_full_cell_analysis.test_full_cell_analysis()

Test FullCellAnalysis computes convex volumes and complexities.

This test builds two synthetic 3×3×3 masks, each containing four voxels arranged in an L-shape. These small shapes have simple convex hulls, making it straightforward to verify correctness of the analysis.

Verifications performed:

• Convex hull simplices and vertex indices are consistent with the voxel geometry.

• Convex hull volumes are computed in physical units by scaling with voxscale.

• Cell complexities are correctly calculated as convex_volume / cell_volume.

• The maximum cell volume is determined correctly from voxel counts.

Asserts:

• cell_volumes match expected floating-point values.

• convex_vertices match the expected vertex indices for each mask.

• convex_volumes match expected floating-point values.

• cell_complexities match expected floating-point values.

• max_cell_volume matches the expected maximum cell volume.

pycroglia.core.tests.unit.test_nearest_pixel.test_nearest_pixel()

Test NearestPixel correctly finds the closest foreground voxel.

Scenario:

• A 5×5×5 binary volume is created with a single voxel set to True at coordinates (z=2, y=3, x=4).

• The search starts at (0,0,0) with isotropic scale=1.0.

Asserts:

• The nearest voxel coordinates returned by compute match the expected location (x=4, y=3, z=2).

• The returned result is equal to a NearestPixelCoordinates instance with these values.

pycroglia.core.tests.unit.test_reorder.test_diagonal_path()

Test reorder_pixel_list correctly orders a 3D diagonal line.

Asserts:

• Voxels are reordered from endpoint (0,0,0) to centroid (4,4,4).

• The resulting path matches [[0,0,0], [1,1,1], …, [4,4,4]].

pycroglia.core.tests.unit.test_reorder.test_simple_line_path()

Test reorder_pixel_list correctly orders a straight vertical line.

Asserts:

• Voxels are reordered from the endpoint at z=0 to the centroid at z=4.

• The resulting order matches [[0,2,2], [1,2,2], …, [4,2,2]].

pycroglia.core.tests.unit.test_segmentation.simple_cells_img()

Returns a 3D binary image with two separate cells.

Returns:

3D binary image.

Return type:

NDArray

pycroglia.core.tests.unit.test_segmentation.test_segment_cell_basic()

Test segment_cell returns correct number of segments for simple input.

Asserts:

The number of returned segments matches the number of cells.

pycroglia.core.tests.unit.test_segmentation.test_segment_cell_small_cell_not_split()

Test segment_cell does not split small cells below cut_off_size.

Asserts:

The returned segment is the same as the input cell.

pycroglia.core.tests.unit.test_segmentation.test_segment_single_cell_split()

Test segment_single_cell splits a large cell into multiple segments.

Asserts:

The number of returned segments matches the number of nuclei.

pycroglia.core.tests.unit.test_segmentation.test_segmentation_config_defaults()

Test SegmentationConfig default values.

Asserts:

The default values are set as expected.

pycroglia.core.tests.unit.test_territorial_volume.test_compute_metrics()

Test compute_metrics returns correct global volume metrics.

Asserts:

All TerritorialVolumeMetrics fields (total_volume_covered, image_cube_volume, empty_volume, covered_percentage) match expected reference values.

pycroglia.core.tests.unit.test_territorial_volume.test_compute_territorial_volume()

Test TerritorialVolume.compute returns correct convex volumes for small masks.

Asserts:

The computed convex hull volumes match expected reference values.

Module contents