tensorflow
/
probability
mirror of https://github.com/tensorflow/probability.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

[numpy] Replace `interpolation=` with `method=` on numpy's `percentile`, `quantile`, `nanpercentile` and `nanquantile` APIs. 

The `interpolation` keyword argument is removed in NumPy 2.4.

PiperOrigin-RevId: 840909914
This commit is contained in:
phawkins 2025-12-05 16:01:09 -08:00 committed by TensorFlower Gardener
parent e4dd81549d
commit 739846f0e7
1 changed files with 19 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
tensorflow_probability/python/stats/quantiles_test.py
View File

@ -440,7 +440,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
x = [1., 5., 3., 2., 4.]
for q in [0, 10, 25, 49.9, 50, 50.01, 90, 95, 100]:
expected_percentile = np.percentile(
x, q=q, interpolation=self._interpolation, axis=0)
x, q=q, method=self._interpolation, axis=0)
pct = quantiles.percentile(
x, q=q, interpolation=self._interpolation, axis=[0])
self.assertAllEqual((), pct.shape)
@ -450,7 +450,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
x = [1., 5., 3., 2., 4.]
q = np.array([0, 10, 25, 49.9, 50, 50.01, 90, 95, 100])
expected_percentile = np.percentile(
x, q=q, interpolation=self._interpolation, axis=0)
x, q=q, method=self._interpolation, axis=0)
pct = quantiles.percentile(
x, q=q, interpolation=self._interpolation, axis=[0])
self.assertAllEqual(q.shape, pct.shape)
@ -460,7 +460,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
x = [1., 5., 3., 2., 4., 5.]
for q in [0, 10, 25, 49.9, 50, 50.01, 90, 95, 100]:
expected_percentile = np.percentile(
x, q=q, interpolation=self._interpolation)
x, q=q, method=self._interpolation)
pct = quantiles.percentile(x, q=q, interpolation=self._interpolation)
self.assertAllEqual((), pct.shape)
self.assertAllClose(expected_percentile, self.evaluate(pct))
@ -469,7 +469,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
x = np.array([[-1., 50., -3.5, 2., -1], [0., 0., 3., 2., 4.]]).T
for q in [0, 10, 25, 49.9, 50, 50.01, 90, 95, 100]:
expected_percentile = np.percentile(
x, q=q, interpolation=self._interpolation, axis=0)
x, q=q, method=self._interpolation, axis=0)
# Get dim 1 with negative and positive indices.
pct_neg_index = quantiles.percentile(
x, q=q, interpolation=self._interpolation, axis=[0])
@ -485,7 +485,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
x = np.array([1., 2., 4., 50.])
q = 10
expected_percentile = np.percentile(
x, q=q, interpolation=self._interpolation, axis=0)
x, q=q, method=self._interpolation, axis=0)
pct = quantiles.percentile(
x, q=q, interpolation=self._interpolation, axis=[0])
self.assertAllClose(expected_percentile, self.evaluate(pct))
@ -494,7 +494,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
x = np.array([[1., 2., 4., 50.], [1., 2., -4., 5.]]).T
for q in [0, 10, 25, 49.9, 50, 50.01, 90, 95, 100]:
expected_percentile = np.percentile(
x, q=q, interpolation=self._interpolation, axis=0)
x, q=q, method=self._interpolation, axis=0)
pct = quantiles.percentile(
x, q=q, interpolation=self._interpolation, axis=[0])
self.assertAllEqual((2,), pct.shape)
@ -504,7 +504,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
x = np.array([[1., 2., 4., 50.], [1., 2., -4., 5.]]).T
for q in [0, 10, 25, 49.9, 50, 50.01, 90, 95, 100]:
expected_percentile = np.percentile(
x, q=q, interpolation=self._interpolation, keepdims=True, axis=0)
x, q=q, method=self._interpolation, keepdims=True, axis=0)
pct = quantiles.percentile(
x, q=q, interpolation=self._interpolation, keepdims=True, axis=[0])
self.assertAllEqual((1, 2), pct.shape)
@ -514,7 +514,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
x = np.random.rand(2, 3, 4, 5)
for axis in [None, 0, 1, -2, (0,), (-1,), (-1, 1), (3, 1), (-3, 0)]:
expected_percentile = np.percentile(
x, q=0.77, interpolation=self._interpolation, axis=axis)
x, q=0.77, method=self._interpolation, axis=axis)
pct = quantiles.percentile(
x, q=0.77, interpolation=self._interpolation, axis=axis)
self.assertAllEqual(expected_percentile.shape, pct.shape)
@ -525,7 +525,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
q = [0.25, 0.75]
for axis in [None, 0, (-1, 1)]:
expected_percentile = np.percentile(
x, q=q, interpolation=self._interpolation, axis=axis)
x, q=q, method=self._interpolation, axis=axis)
pct = quantiles.percentile(
x, q=q, interpolation=self._interpolation, axis=axis)
self.assertAllEqual(expected_percentile.shape, pct.shape)
@ -536,7 +536,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
q = [0.25, 0.75]
for axis in [None, 0, (-1, 1)]:
expected_percentile = np.percentile(
x, q=q, interpolation=self._interpolation, axis=axis, keepdims=True)
x, q=q, method=self._interpolation, axis=axis, keepdims=True)
pct = quantiles.percentile(
x, q=q, interpolation=self._interpolation, axis=axis, keepdims=True)
self.assertAllEqual(expected_percentile.shape, pct.shape)
@ -548,7 +548,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
expected_percentile = np.percentile(
x,
q=0.77,
interpolation=self._interpolation,
method=self._interpolation,
axis=axis,
keepdims=True)
pct = quantiles.percentile(
@ -565,7 +565,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
x_ph = tf1.placeholder_with_default(x, shape=[None, None, None, None])
for axis in [None, 0, 1, -2, (0,), (-1,), (-1, 1), (3, 1), (-3, 0)]:
expected_percentile = np.percentile(
x, q=0.77, interpolation=self._interpolation, axis=axis)
x, q=0.77, method=self._interpolation, axis=axis)
pct = quantiles.percentile(
x_ph, q=0.77, interpolation=self._interpolation, axis=axis)
self.assertAllClose(expected_percentile, self.evaluate(pct))
@ -577,7 +577,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
expected_percentile = np.percentile(
x,
q=0.77,
interpolation=self._interpolation,
method=self._interpolation,
axis=axis,
keepdims=True)
pct = quantiles.percentile(
@ -595,7 +595,7 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
x = [1, 5, 3, 2, 4]
for q in [0, 10, 25, 49.9, 50, 50.01, 90, 95, 100]:
expected_percentile = np.percentile(
x, q=q, interpolation=self._interpolation)
x, q=q, method=self._interpolation)
pct = quantiles.percentile(x, q=q, interpolation=self._interpolation)
self.assertEqual(tf.int32, pct.dtype)
self.assertAllEqual((), pct.shape)
@ -610,13 +610,13 @@ class PercentileTestWithLowerInterpolation(test_util.TestCase):
# Test each percentile individually
for q in qs:
expected_percentile = np.percentile(
xs, q=q, interpolation=self._interpolation)
xs, q=q, method=self._interpolation)
self.assertTrue(np.isnan(expected_percentile))
pct = quantiles.percentile(xs, q=q, interpolation=self._interpolation)
self.assertTrue(self.evaluate(tf.math.is_nan(pct)))
# Test vector percentile as well
expected_percentiles = np.percentile(
xs, q=qs, interpolation=self._interpolation)
xs, q=qs, method=self._interpolation)
pcts = quantiles.percentile(xs, q=qs, interpolation=self._interpolation)
self.assertAllEqual(expected_percentiles, pcts)
@ -793,7 +793,7 @@ class PercentileTestWithNearestInterpolation(test_util.TestCase):
x = [1., 5., 3., 2., 4.]
for q in [0, 10.1, 25.1, 49.9, 50.1, 50.01, 89, 100]:
expected_percentile = np.percentile(
x, q=q, interpolation=self._interpolation)
x, q=q, method=self._interpolation)
pct = quantiles.percentile(x, q=q, interpolation=self._interpolation)
self.assertAllEqual((), pct.shape)
self.assertAllClose(expected_percentile, self.evaluate(pct))
@ -802,7 +802,7 @@ class PercentileTestWithNearestInterpolation(test_util.TestCase):
x = [1., 5., 3., 2., 4., 5.]
for q in [0, 10.1, 25.1, 49.9, 50.1, 50.01, 89, 100]:
expected_percentile = np.percentile(
x, q=q, interpolation=self._interpolation)
x, q=q, method=self._interpolation)
pct = quantiles.percentile(x, q=q, interpolation=self._interpolation)
self.assertAllEqual((), pct.shape)
self.assertAllClose(expected_percentile, self.evaluate(pct))
@ -811,7 +811,7 @@ class PercentileTestWithNearestInterpolation(test_util.TestCase):
x = [[10.0, 7.0, 4.0], [3.0, 2.0, 1.0]]
for q in [0, 10.1, 25.1, 49.9, 50.0, 50.1, 50.01, 89, 100]:
expected_percentile = np.percentile(
x, q=q, interpolation=self._interpolation)
x, q=q, method=self._interpolation)
pct = quantiles.percentile(x, q=q, interpolation=self._interpolation)
self.assertAllEqual((), pct.shape)
self.assertAllClose(expected_percentile, self.evaluate(pct))