def test_spearman_nan_result():
"""
Confirms that any nan value on the way in
will result in a nan value returned.
"""
x = np.array([1, 2, 3, 4], dtype='float')
y = np.array([2, 3, 4, 3], dtype='float')
assert spearman(x, y) == approx(0.8)
x[0] = np.nan
assert spearman(x, y) == approx(0.4)
x[0] = 1
y[0] = np.nan
assert spearman(x, y) == approx(-0.4)
y[0] = 2
assert spearman(x, y) == approx(0.8)
def test_spearman_rgs():
"""
Tests the Spearman rank correlation using
the Royal Geographical Society example pdf in
literature/OASpearmansRankExcelGuidePDF.pdf
"""
width = np.array([0, 50, 150, 200, 250, 300, 350, 400, 450, 500])
depth = np.array([0, 10, 28, 42, 59, 51, 73, 85, 104, 96])
assert spearman(width, depth) == approx(0.9757, abs=1e-4)
A = np.stack([width, depth]).T
assert spearman_pairwise(A).diagonal(1) == approx(0.9757, abs=1e-4)
def test_spearman_basic_sanity():
"""
Tests the Spearman rank correlation using
the example on wikipedia:
https://en.wikipedia.org/wiki/Spearman%27s_rank_correlation_coefficient
"""
iq = np.array([86, 97, 99, 100, 101, 103, 106, 110, 112, 113])
tv = np.array([0, 20, 28, 27, 50, 29, 7, 17, 6, 12])
result = spearman(iq, tv)
assert result == approx(-0.1757575, abs=1e-7)
A = np.stack([iq, tv]).T
assert spearman_pairwise(A).diagonal(1) == approx(-0.1757575, abs=1e-7)