101 lines
4.5 KiB
Python
101 lines
4.5 KiB
Python
import numpy as np
|
|
import pytest
|
|
from skimage.measure import (intersection_coeff, manders_coloc_coeff,
|
|
manders_overlap_coeff, pearson_corr_coeff)
|
|
|
|
|
|
def test_invalid_input():
|
|
# images are not same size
|
|
img1 = np.array([[i + j for j in range(4)] for i in range(4)])
|
|
img2 = np.ones((3, 5, 6))
|
|
mask = np.array([[i <= 1 for i in range(5)] for _ in range(5)])
|
|
non_binary_mask = np.array([[2 for __ in range(4)] for _ in range(4)])
|
|
|
|
with pytest.raises(ValueError, match=". must have the same dimensions"):
|
|
pearson_corr_coeff(img1, img1, mask)
|
|
with pytest.raises(ValueError, match=". must have the same dimensions"):
|
|
pearson_corr_coeff(img1, img2)
|
|
with pytest.raises(ValueError, match=". must have the same dimensions"):
|
|
pearson_corr_coeff(img1, img1, mask)
|
|
with pytest.raises(ValueError, match=". array is not of dtype boolean"):
|
|
pearson_corr_coeff(img1, img1, non_binary_mask)
|
|
with pytest.raises(ValueError, match=". must have the same dimensions"):
|
|
manders_coloc_coeff(img1, mask)
|
|
with pytest.raises(ValueError, match=". array is not of dtype boolean"):
|
|
manders_coloc_coeff(img1, non_binary_mask)
|
|
with pytest.raises(ValueError, match=". must have the same dimensions"):
|
|
manders_coloc_coeff(img1, img1 > 0, mask)
|
|
with pytest.raises(ValueError, match=". array is not of dtype boolean"):
|
|
manders_coloc_coeff(img1, img1 > 0, non_binary_mask)
|
|
with pytest.raises(ValueError, match=". must have the same dimensions"):
|
|
manders_overlap_coeff(img1, img1, mask)
|
|
with pytest.raises(ValueError, match=". must have the same dimensions"):
|
|
manders_overlap_coeff(img1, img2)
|
|
with pytest.raises(ValueError, match=". must have the same dimensions"):
|
|
manders_overlap_coeff(img1, img1, mask)
|
|
with pytest.raises(ValueError, match=". array is not of dtype boolean"):
|
|
manders_overlap_coeff(img1, img1, non_binary_mask)
|
|
with pytest.raises(ValueError, match=". must have the same dimensions"):
|
|
intersection_coeff(img1 > 2, img2 > 1, mask)
|
|
with pytest.raises(ValueError, match=". array is not of dtype boolean"):
|
|
intersection_coeff(img1, img2)
|
|
with pytest.raises(ValueError, match=". must have the same dimensions"):
|
|
intersection_coeff(img1 > 2, img1 > 1, mask)
|
|
with pytest.raises(ValueError, match=". array is not of dtype boolean"):
|
|
intersection_coeff(img1 > 2, img1 > 1, non_binary_mask)
|
|
|
|
|
|
def test_pcc():
|
|
# simple example
|
|
img1 = np.array([[i + j for j in range(4)] for i in range(4)])
|
|
assert pearson_corr_coeff(img1, img1) == (1.0, 0.0)
|
|
|
|
img2 = np.where(img1 <= 2, 0, img1)
|
|
np.testing.assert_almost_equal(pearson_corr_coeff(img1, img2), (0.944911182523068, 3.5667540654536515e-08))
|
|
|
|
# change background of roi and see if values are same
|
|
roi = np.where(img1 <= 2, 0, 1)
|
|
np.testing.assert_almost_equal(pearson_corr_coeff(img1, img1, roi), pearson_corr_coeff(img1, img2, roi))
|
|
|
|
|
|
def test_mcc():
|
|
img1 = np.array([[j for j in range(4)] for i in range(4)])
|
|
mask = np.array([[i <= 1 for j in range(4)]for i in range(4)])
|
|
assert manders_coloc_coeff(img1, mask) == 0.5
|
|
|
|
# test negative values
|
|
img_negativeint = np.where(img1 == 1, -1, img1)
|
|
img_negativefloat = img_negativeint / 2.0
|
|
with pytest.raises(ValueError):
|
|
manders_coloc_coeff(img_negativeint, mask)
|
|
with pytest.raises(ValueError):
|
|
manders_coloc_coeff(img_negativefloat, mask)
|
|
|
|
|
|
def test_moc():
|
|
img1 = np.ones((4, 4))
|
|
img2 = 2 * np.ones((4, 4))
|
|
assert manders_overlap_coeff(img1, img2) == 1
|
|
|
|
# test negative values
|
|
img_negativeint = np.where(img1 == 1, -1, img1)
|
|
img_negativefloat = img_negativeint / 2.0
|
|
with pytest.raises(ValueError):
|
|
manders_overlap_coeff(img_negativeint, img2)
|
|
with pytest.raises(ValueError):
|
|
manders_overlap_coeff(img1, img_negativeint)
|
|
with pytest.raises(ValueError):
|
|
manders_overlap_coeff(img_negativefloat, img2)
|
|
with pytest.raises(ValueError):
|
|
manders_overlap_coeff(img1, img_negativefloat)
|
|
with pytest.raises(ValueError):
|
|
manders_overlap_coeff(img_negativefloat, img_negativefloat)
|
|
|
|
|
|
def test_intersection_coefficient():
|
|
img1_mask = np.array([[j <= 1 for j in range(4)] for i in range(4)])
|
|
img2_mask = np.array([[i <= 1 for j in range(4)] for i in range(4)])
|
|
img3_mask = np.array([[1 for j in range(4)] for i in range(4)])
|
|
assert intersection_coeff(img1_mask, img2_mask) == 0.5
|
|
assert intersection_coeff(img1_mask, img3_mask) == 1
|