def test_call_function(make_data):
data, expected = make_data()
t = ClinicalKernelTransform(fit_once=True)
t.prepare(data)
mat = t(t.X_fit_, t.X_fit_)
assert_array_almost_equal(expected, mat, 4)
def test_compare_clinical_kernel(self):
x_full, y = load_whas500()
trans = ClinicalKernelTransform()
trans.fit(x_full)
kpca = KernelPCA(kernel=trans.pairwise_kernel, copy_X=True)
xt = kpca.fit_transform(self.x)
nrsvm = FastSurvivalSVM(optimizer='rbtree',
tol=1e-8,
max_iter=500,
random_state=0)
nrsvm.fit(xt, y)
rsvm = FastKernelSurvivalSVM(optimizer='rbtree',
kernel=trans.pairwise_kernel,
tol=1e-8,
max_iter=500,
random_state=0)
rsvm.fit(self.x.values, y)
pred_nrsvm = nrsvm.predict(kpca.transform(self.x))
pred_rsvm = rsvm.predict(self.x.values)
self.assertEqual(len(pred_nrsvm), len(pred_rsvm))
c1 = concordance_index_censored(y['fstat'], y['lenfol'], pred_nrsvm)
c2 = concordance_index_censored(y['fstat'], y['lenfol'], pred_rsvm)
self.assertAlmostEqual(c1[0], c2[0])
self.assertTupleEqual(c1[1:], c2[1:])
def test_fit_custom_kernel(make_whas500):
whas500 = make_whas500(with_mean=False, with_std=False, to_numeric=True)
alphas = numpy.exp(numpy.linspace(numpy.log(0.001), numpy.log(0.5), 5))
svm_grid = ParameterGrid({"alpha": alphas})
transform = ClinicalKernelTransform(fit_once=True)
transform.prepare(whas500.x_data_frame)
base_estimators = []
for i, params in enumerate(svm_grid):
model = FastSurvivalSVM(max_iter=100, random_state=0, **params)
base_estimators.append(("svm_linear_%d" % i, model))
for i, params in enumerate(svm_grid):
model = FastKernelSurvivalSVM(kernel=transform.pairwise_kernel, max_iter=45, tol=1e-5,
random_state=0, **params)
base_estimators.append(("svm_kernel_%d" % i, model))
cv = KFold(n_splits=3, shuffle=True, random_state=0)
meta = EnsembleSelection(base_estimators, n_estimators=0.4, scorer=score_cindex, cv=cv, n_jobs=4)
meta.fit(whas500.x, whas500.y)
assert len(meta) == 10
assert meta.scores_.shape == (10,)
p = meta.predict(whas500.x)
assert_cindex_almost_equal(whas500.y['fstat'], whas500.y['lenfol'], p,
(0.7978084, 59938, 15178, 33, 119))
def test_compare_clinical_kernel(make_whas500):
whas500 = make_whas500(to_numeric=True)
trans = ClinicalKernelTransform()
trans.fit(whas500.x_data_frame)
kpca = KernelPCA(kernel=trans.pairwise_kernel, copy_X=True)
xt = kpca.fit_transform(whas500.x)
nrsvm = FastSurvivalSVM(optimizer='rbtree',
tol=1e-8,
max_iter=500,
random_state=0)
nrsvm.fit(xt, whas500.y)
rsvm = FastKernelSurvivalSVM(optimizer='rbtree',
kernel=trans.pairwise_kernel,
tol=1e-8,
max_iter=500,
random_state=0)
rsvm.fit(whas500.x, whas500.y)
pred_nrsvm = nrsvm.predict(kpca.transform(whas500.x))
pred_rsvm = rsvm.predict(whas500.x)
assert len(pred_nrsvm) == len(pred_rsvm)
expected_cindex = concordance_index_censored(whas500.y['fstat'],
whas500.y['lenfol'],
pred_nrsvm)
assert_cindex_almost_equal(whas500.y['fstat'], whas500.y['lenfol'],
pred_rsvm, expected_cindex)
def test_kernel_transform_feature_mismatch(self):
t = ClinicalKernelTransform()
t.fit(self.data)
self.assertRaisesRegex(ValueError,
'expected array with 4 features, but got 17',
t.transform, numpy.zeros((2, 17), dtype=float))
def test_fit_custom_kernel(self):
alphas = numpy.exp(numpy.linspace(numpy.log(0.001), numpy.log(0.5), 5))
svm_grid = ParameterGrid({"alpha": alphas})
transform = ClinicalKernelTransform(fit_once=True)
transform.prepare(self.x)
base_estimators = []
for i, params in enumerate(svm_grid):
model = FastSurvivalSVM(max_iter=100, random_state=0, **params)
base_estimators.append(("svm_linear_%d" % i, model))
for i, params in enumerate(svm_grid):
model = FastKernelSurvivalSVM(kernel=transform.pairwise_kernel, max_iter=45, tol=1e-5,
random_state=0, **params)
base_estimators.append(("svm_kernel_%d" % i, model))
cv = KFold(n_splits=3, shuffle=True, random_state=0)
meta = EnsembleSelection(base_estimators, n_estimators=0.4, scorer=score_cindex, cv=cv, n_jobs=4)
meta.fit(self.x.values, self.y)
self.assertEqual(len(meta), 10)
self.assertTupleEqual(meta.scores_.shape, (10,))
p = meta.predict(self.x.values)
score = concordance_index_censored(self.y['fstat'], self.y['lenfol'], p)
expected_score = numpy.array([0.7978084, 59938, 15178, 33, 119])
assert_array_almost_equal(score, expected_score)
def test_call_function(self):
t = ClinicalKernelTransform(fit_once=True)
t.prepare(self.data)
mat = t(t.X_fit_, t.X_fit_)
expected = _get_expected_matrix()
assert_array_almost_equal(expected, mat, 4)
def test_kernel_transform(make_data):
data, expected = make_data()
t = ClinicalKernelTransform()
t.fit(data)
mat = t.transform(t.X_fit_)
assert_array_almost_equal(expected, mat, 4)
def test_pairwise_feature_mismatch(self):
t = ClinicalKernelTransform()
t.fit(self.data)
self.assertRaisesRegex(ValueError, r'Incompatible dimension for X and Y matrices: '
r'X.shape\[1\] == 4 while Y.shape\[1\] == 17',
pairwise_kernels, t.X_fit_, numpy.zeros((2, 17), dtype=float),
metric=t.pairwise_kernel, n_jobs=1)
def test_pairwise_x_and_y_error_shape(self):
t = ClinicalKernelTransform()
t.fit(self.data)
self.assertRaisesRegex(ValueError,
"X and Y have different number of features",
t.pairwise_kernel, self.data.iloc[0, :],
self.data.iloc[1, :2])
def test_kernel_transform(self):
t = ClinicalKernelTransform()
t.fit(self.data)
mat = t.transform(t.X_fit_)
expected = _get_expected_matrix()
assert_array_almost_equal(expected, mat, 4)
def test_pairwise_no_nominal(make_data):
data, expected = make_data(with_nominal=False)
t = ClinicalKernelTransform()
t.fit(data)
mat = pairwise_kernels(t.X_fit_[:3, :], t.X_fit_[3:, :],
metric=t.pairwise_kernel, n_jobs=1)
assert_array_almost_equal(expected[:3:, 3:], mat, 4)
def test_prepare(self):
t = ClinicalKernelTransform(fit_once=True)
t.prepare(self.data)
copy = clone(t).fit(t.X_fit_)
mat = copy.transform(t.X_fit_[:4, :])
expected = _get_expected_matrix()
assert_array_almost_equal(expected[:4, :], mat, 4)
def test_prepare_error_dtype():
t = ClinicalKernelTransform(fit_once=True)
data = pandas.DataFrame({"age": [12, 61, 18, 21, 57, 17],
"date": numpy.array(
["2016-01-01", "1954-06-30", "1999-03-01", "2005-02-25", "2112-12-31",
"1731-09-16"], dtype='datetime64')})
with pytest.raises(TypeError, match=r'unsupported dtype: dtype\(.+\)'):
t.prepare(data)
def test_pairwise_feature_mismatch(make_data):
data, _ = make_data()
t = ClinicalKernelTransform()
t.fit(data)
with pytest.raises(ValueError, match=r'Incompatible dimension for X and Y matrices: '
r'X.shape\[1\] == 4 while Y.shape\[1\] == 17'):
pairwise_kernels(t.X_fit_, numpy.zeros((2, 17), dtype=float),
metric=t.pairwise_kernel, n_jobs=1)
def test_pairwise_no_nominal(self):
t = ClinicalKernelTransform()
t.fit(self.data.drop('metastasis', axis=1))
mat = pairwise_kernels(t.X_fit_[:3, :], t.X_fit_[3:, :],
metric=t.pairwise_kernel, n_jobs=1)
expected = _get_expected_matrix(with_nominal=False)
assert_array_almost_equal(expected[:3:, 3:], mat, 4)
def test_pairwise(make_data):
data, expected = make_data()
t = ClinicalKernelTransform()
t.fit(data)
mat = pairwise_kernels(t.X_fit_, t.X_fit_,
metric=t.pairwise_kernel, n_jobs=1)
assert_array_almost_equal(expected, mat, 4)
def test_call_function_x_and_y(self):
t = ClinicalKernelTransform(fit_once=True)
t.prepare(self.data)
mat = t(t.X_fit_[:3, :], t.X_fit_[3:, :])
m = _get_expected_matrix()
expected = m[:3:, 3:]
assert_array_almost_equal(expected, mat, 4)
def test_call_function_x_and_y(make_data):
data, m = make_data()
t = ClinicalKernelTransform(fit_once=True)
t.prepare(data)
mat = t(t.X_fit_[:3, :], t.X_fit_[3:, :])
expected = m[:3:, 3:]
assert_array_almost_equal(expected, mat, 4)
def test_prepare(make_data):
data, expected = make_data()
t = ClinicalKernelTransform(fit_once=True)
t.prepare(data)
copy = clone(t).fit(t.X_fit_)
mat = copy.transform(t.X_fit_[:4, :])
assert_array_almost_equal(expected[:4, :], mat, 4)
def test_pairwise(self):
t = ClinicalKernelTransform()
t.fit(self.data)
mat = pairwise_kernels(t.X_fit_, t.X_fit_,
metric=t.pairwise_kernel, n_jobs=1)
expected = _get_expected_matrix()
assert_array_almost_equal(expected, mat, 4)
def test_prepare_error_fit_once(make_data):
data = make_data()
t = ClinicalKernelTransform(fit_once=False)
with pytest.raises(
ValueError,
match=
"prepare can only be used if fit_once parameter is set to True"
):
t.prepare(data)
def test_pairwise_x_and_y(make_data):
data, m = make_data()
t = ClinicalKernelTransform()
t.fit(data)
mat = pairwise_kernels(t.X_fit_[:3, :], t.X_fit_[3:, :],
metric=t.pairwise_kernel, n_jobs=1)
expected = m[:3:, 3:]
assert_array_almost_equal(expected, mat, 4)
def test_pairwise_x_and_y(self):
t = ClinicalKernelTransform()
t.fit(self.data)
mat = pairwise_kernels(t.X_fit_[:3, :], t.X_fit_[3:, :],
metric=t.pairwise_kernel, n_jobs=1)
m = _get_expected_matrix()
expected = m[:3:, 3:]
assert_array_almost_equal(expected, mat, 4)
def test_fit_and_predict_clinical_kernel(make_whas500):
whas500 = make_whas500(to_numeric=True)
trans = ClinicalKernelTransform()
trans.fit(whas500.x_data_frame)
ssvm = FastKernelSurvivalSVM(optimizer="rbtree", kernel=trans.pairwise_kernel,
tol=7e-7, max_iter=100, random_state=0)
ssvm.fit(whas500.x, whas500.y)
assert not ssvm._pairwise
assert whas500.x.shape[0] == ssvm.coef_.shape[0]
c = ssvm.score(whas500.x, whas500.y)
assert c >= 0.854
def test_prepare_error_fit_once(self):
t = ClinicalKernelTransform(fit_once=False)
self.assertRaisesRegex(
ValueError,
"prepare can only be used if fit_once parameter is set to True",
t.prepare, self.data)
def test_fit_and_predict_clinical_kernel(self):
x_full, y = load_whas500()
trans = ClinicalKernelTransform()
trans.fit(x_full)
x = self.x
ssvm = FastKernelSurvivalSVM(optimizer="rbtree", kernel=trans.pairwise_kernel,
tol=7e-7, max_iter=100, random_state=0)
ssvm.fit(x.values, y)
self.assertFalse(ssvm._pairwise)
self.assertEquals(x.shape[0], ssvm.coef_.shape[0])
c = ssvm.score(x.values, y)
self.assertGreaterEqual(c, 0.854)
def test_fit_error_ndim(self):
t = ClinicalKernelTransform()
self.assertRaisesRegex(ValueError, "expected 2d array, but got 1",
t.fit, numpy.random.randn(31))
self.assertRaisesRegex(ValueError, "expected 2d array, but got 3",
t.fit, numpy.random.randn(31, 20, 2))
def test_prepare_error_dtype(self):
t = ClinicalKernelTransform(fit_once=True)
data = pandas.DataFrame({"age": [12, 61, 18, 21, 57, 17],
"date": numpy.array(
["2016-01-01", "1954-06-30", "1999-03-01", "2005-02-25", "2112-12-31",
"1731-09-16"], dtype='datetime64')})
self.assertRaisesRegex(TypeError, r'unsupported dtype: dtype\(.+\)',
t.prepare, data)
def test_fit_and_predict_clinical_kernel(self):
x_full, y = load_whas500()
trans = ClinicalKernelTransform()
trans.fit(x_full)
x = encode_categorical(standardize(x_full))
ssvm = FastKernelSurvivalSVM(optimizer="rbtree",
kernel=trans.pairwise_kernel,
max_iter=100,
random_state=0)
ssvm.fit(x.values, y)
self.assertFalse(ssvm._pairwise)
self.assertEquals(x.shape[0], ssvm.coef_.shape[0])
c = ssvm.score(x.values, y)
self.assertLessEqual(abs(0.83699051218246412 - c), 1e-3)
