Python warnings.catch_warnings函数代码示例

  def testWarnings(self):
    # Smaller than the threshold: no warning.
    c_sparse = ops.IndexedSlices(array_ops.placeholder(dtypes.float32),
                                 array_ops.placeholder(dtypes.int32),
                                 constant([4, 4, 4, 4]))
    with warnings.catch_warnings(record=True) as w:
      math_ops.mul(c_sparse, 1.0)
    self.assertEqual(0, len(w))

    # Greater than or equal to the threshold: warning.
    c_sparse = ops.IndexedSlices(array_ops.placeholder(dtypes.float32),
                                 array_ops.placeholder(dtypes.int32),
                                 constant([100, 100, 100, 100]))
    with warnings.catch_warnings(record=True) as w:
      math_ops.mul(c_sparse, 1.0)
    self.assertEqual(1, len(w))
    self.assertTrue(
        "with 100000000 elements. This may consume a large amount of memory."
        in str(w[0].message))

    # Unknown dense shape: warning.
    c_sparse = ops.IndexedSlices(array_ops.placeholder(dtypes.float32),
                                 array_ops.placeholder(dtypes.int32),
                                 array_ops.placeholder(dtypes.int32))
    with warnings.catch_warnings(record=True) as w:
      math_ops.mul(c_sparse, 1.0)
    self.assertEqual(1, len(w))
    self.assertTrue(
        "of unknown shape. This may consume a large amount of memory."
        in str(w[0].message))

    def test_long_cache_keys_shortened(self):
        cache_settings = {
            'default': {
                'BACKEND': 'django.core.cache.backends.filebased.FileBasedCache',
                'LOCATION': os.path.join(TOP_DIR, 'test.cache'),
            }
        }
        long_key_string = "X" * 251

        with override_settings(CACHES=cache_settings):
            with warnings.catch_warnings(record=True) as w:
                warnings.simplefilter("always")
                # memcached limits key length to 250
                cache.set(long_key_string, "hello cached world")

                self.assertEqual(len(w), 1)
                self.assertIsInstance(w[0].message, CacheKeyWarning)

        # Activate optional cache key length checker
        cache_settings['default']['KEY_FUNCTION'] = 'mainsite.utils.filter_cache_key'

        with override_settings(CACHES=cache_settings):
            with warnings.catch_warnings(record=True) as w:
                warnings.simplefilter("always")
                # memcached limits key length to 250
                cache.set(long_key_string, "hello cached world")

                self.assertEqual(len(w), 0)

                retrieved = cache.get(long_key_string)

                self.assertEqual(retrieved, "hello cached world")

def test_calculate_chpi_positions():
    """Test calculation of cHPI positions
    """
    trans, rot, t = head_pos_to_trans_rot_t(read_head_pos(pos_fname))
    with warnings.catch_warnings(record=True):
        raw = Raw(chpi_fif_fname, allow_maxshield=True, preload=True)
    t -= raw.first_samp / raw.info['sfreq']
    quats = _calculate_chpi_positions(raw, verbose='debug')
    trans_est, rot_est, t_est = head_pos_to_trans_rot_t(quats)
    _compare_positions((trans, rot, t), (trans_est, rot_est, t_est), 0.003)

    # degenerate conditions
    raw_no_chpi = Raw(test_fif_fname)
    assert_raises(RuntimeError, _calculate_chpi_positions, raw_no_chpi)
    raw_bad = raw.copy()
    for d in raw_bad.info['dig']:
        if d['kind'] == FIFF.FIFFV_POINT_HPI:
            d['coord_frame'] = 999
            break
    assert_raises(RuntimeError, _calculate_chpi_positions, raw_bad)
    raw_bad = raw.copy()
    for d in raw_bad.info['dig']:
        if d['kind'] == FIFF.FIFFV_POINT_HPI:
            d['r'] = np.ones(3)
    raw_bad.crop(0, 1., copy=False)
    with warnings.catch_warnings(record=True):  # bad pos
        with catch_logging() as log_file:
            _calculate_chpi_positions(raw_bad, verbose=True)
    # ignore HPI info header and [done] footer
    for line in log_file.getvalue().strip().split('\n')[4:-1]:
        assert_true('0/5 good' in line)

def test_misspecifications():
    # Tests for model specification and misspecification exceptions
    endog = np.arange(20).reshape(10,2)

    # Bad trend specification
    assert_raises(ValueError, varmax.VARMAX, endog, order=(1,0), trend='')

    # Bad error_cov_type specification
    assert_raises(ValueError, varmax.VARMAX, endog, order=(1,0), error_cov_type='')

    # Bad order specification
    assert_raises(ValueError, varmax.VARMAX, endog, order=(0,0))

    with warnings.catch_warnings(record=True) as w:
        warnings.simplefilter('always')
        varmax.VARMAX(endog, order=(1,1))

    # Warning with VARMA specification
    with warnings.catch_warnings(record=True) as w:
        warnings.simplefilter('always')

        varmax.VARMAX(endog, order=(1,1))

        message = ('Estimation of VARMA(p,q) models is not generically robust,'
                   ' due especially to identification issues.')
        assert_equal(str(w[0].message), message)
    warnings.resetwarnings()

    def test_popen(self):
        mswindows = (sys.platform == "win32")

        if mswindows:
            command = '"{}" -c "print(\'Hello\')"'.format(sys.executable)
        else:
            command = "'{}' -c 'print(\"Hello\")'".format(sys.executable)
        with warnings.catch_warnings():
            warnings.simplefilter("ignore", DeprecationWarning)
            with platform.popen(command) as stdout:
                hello = stdout.read().strip()
                stdout.close()
                self.assertEqual(hello, "Hello")

        data = 'plop'
        if mswindows:
            command = '"{}" -c "import sys; data=sys.stdin.read(); exit(len(data))"'
        else:
            command = "'{}' -c 'import sys; data=sys.stdin.read(); exit(len(data))'"
        command = command.format(sys.executable)
        with warnings.catch_warnings():
            warnings.simplefilter("ignore", DeprecationWarning)
            with platform.popen(command, 'w') as stdin:
                stdout = stdin.write(data)
                ret = stdin.close()
                self.assertIsNotNone(ret)
                if os.name == 'nt':
                    returncode = ret
                else:
                    returncode = ret >> 8
                self.assertEqual(returncode, len(data))

def test_read_epochs_bad_events():
    """Test epochs when events are at the beginning or the end of the file
    """
    # Event at the beginning
    epochs = Epochs(
        raw, np.array([[raw.first_samp, 0, event_id]]), event_id, tmin, tmax, picks=picks, baseline=(None, 0)
    )
    with warnings.catch_warnings(record=True):
        evoked = epochs.average()

    epochs = Epochs(
        raw, np.array([[raw.first_samp, 0, event_id]]), event_id, tmin, tmax, picks=picks, baseline=(None, 0)
    )
    epochs.drop_bad_epochs()
    with warnings.catch_warnings(record=True):
        evoked = epochs.average()

    # Event at the end
    epochs = Epochs(
        raw, np.array([[raw.last_samp, 0, event_id]]), event_id, tmin, tmax, picks=picks, baseline=(None, 0)
    )

    with warnings.catch_warnings(record=True):
        evoked = epochs.average()
        assert evoked
    warnings.resetwarnings()

def test_evoked_io_from_epochs():
    """Test IO of evoked data made from epochs
    """
    # offset our tmin so we don't get exactly a zero value when decimating
    with warnings.catch_warnings(record=True) as w:
        warnings.simplefilter("always")
        epochs = Epochs(raw, events[:4], event_id, tmin + 0.011, tmax, picks=picks, baseline=(None, 0), decim=5)
    assert_true(len(w) == 1)
    evoked = epochs.average()
    evoked.save(op.join(tempdir, "evoked-ave.fif"))
    evoked2 = read_evokeds(op.join(tempdir, "evoked-ave.fif"))[0]
    assert_allclose(evoked.data, evoked2.data, rtol=1e-4, atol=1e-20)
    assert_allclose(evoked.times, evoked2.times, rtol=1e-4, atol=1 / evoked.info["sfreq"])

    # now let's do one with negative time
    with warnings.catch_warnings(record=True) as w:
        warnings.simplefilter("always")
        epochs = Epochs(raw, events[:4], event_id, 0.1, tmax, picks=picks, baseline=(0.1, 0.2), decim=5)
    evoked = epochs.average()
    evoked.save(op.join(tempdir, "evoked-ave.fif"))
    evoked2 = read_evokeds(op.join(tempdir, "evoked-ave.fif"))[0]
    assert_allclose(evoked.data, evoked2.data, rtol=1e-4, atol=1e-20)
    assert_allclose(evoked.times, evoked2.times, rtol=1e-4, atol=1e-20)

    # should be equivalent to a cropped original
    with warnings.catch_warnings(record=True) as w:
        warnings.simplefilter("always")
        epochs = Epochs(raw, events[:4], event_id, -0.2, tmax, picks=picks, baseline=(0.1, 0.2), decim=5)
    evoked = epochs.average()
    evoked.crop(0.099, None)
    assert_allclose(evoked.data, evoked2.data, rtol=1e-4, atol=1e-20)
    assert_allclose(evoked.times, evoked2.times, rtol=1e-4, atol=1e-20)

def _check_predict_proba(clf, X, y):
    proba = clf.predict_proba(X)
    with warnings.catch_warnings():
        warnings.simplefilter("ignore")
        # We know that we can have division by zero
        log_proba = clf.predict_log_proba(X)

    y = np.atleast_1d(y)
    if y.ndim == 1:
        y = np.reshape(y, (-1, 1))

    n_outputs = y.shape[1]
    n_samples = len(X)

    if n_outputs == 1:
        proba = [proba]
        log_proba = [log_proba]

    for k in xrange(n_outputs):
        assert_equal(proba[k].shape[0], n_samples)
        assert_equal(proba[k].shape[1], len(np.unique(y[:, k])))
        assert_array_equal(proba[k].sum(axis=1), np.ones(len(X)))
        with warnings.catch_warnings():
            warnings.simplefilter("ignore")
            # We know that we can have division by zero
            assert_array_equal(np.log(proba[k]), log_proba[k])

    def _check_roundtrip(self, frame):
        _skip_if_no_MySQLdb()
        drop_sql = "DROP TABLE IF EXISTS test_table"
        cur = self.db.cursor()
        with warnings.catch_warnings():
            warnings.filterwarnings("ignore", "Unknown table.*")
            cur.execute(drop_sql)
        sql.write_frame(frame, name='test_table', con=self.db, flavor='mysql')
        result = sql.read_frame("select * from test_table", self.db)

        # HACK! Change this once indexes are handled properly.
        result.index = frame.index
        result.index.name = frame.index.name

        expected = frame
        tm.assert_frame_equal(result, expected)

        frame['txt'] = ['a'] * len(frame)
        frame2 = frame.copy()
        index = Index(lrange(len(frame2))) + 10
        frame2['Idx'] = index
        drop_sql = "DROP TABLE IF EXISTS test_table2"
        cur = self.db.cursor()
        with warnings.catch_warnings():
            warnings.filterwarnings("ignore", "Unknown table.*")
            cur.execute(drop_sql)
        sql.write_frame(frame2, name='test_table2', con=self.db, flavor='mysql')
        result = sql.read_frame("select * from test_table2", self.db,
                                index_col='Idx')
        expected = frame.copy()

        # HACK! Change this once indexes are handled properly.
        expected.index = index
        expected.index.names = result.index.names
        tm.assert_frame_equal(expected, result)

def test_class_weight_classifiers():
    # test that class_weight works and that the semantics are consistent
    classifiers = all_estimators(type_filter="classifier")

    with warnings.catch_warnings(record=True):
        classifiers = [c for c in classifiers if "class_weight" in c[1]().get_params().keys()]

    for n_centers in [2, 3]:
        # create a very noisy dataset
        X, y = make_blobs(centers=n_centers, random_state=0, cluster_std=20)
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5, random_state=0)
        for name, Classifier in classifiers:
            if name == "NuSVC":
                # the sparse version has a parameter that doesn't do anything
                continue
            if name.endswith("NB"):
                # NaiveBayes classifiers have a somewhat different interface.
                # FIXME SOON!
                continue
            if n_centers == 2:
                class_weight = {0: 1000, 1: 0.0001}
            else:
                class_weight = {0: 1000, 1: 0.0001, 2: 0.0001}

            with warnings.catch_warnings(record=True):
                classifier = Classifier(class_weight=class_weight)
            if hasattr(classifier, "n_iter"):
                classifier.set_params(n_iter=100)

            set_random_state(classifier)
            classifier.fit(X_train, y_train)
            y_pred = classifier.predict(X_test)
            assert_greater(np.mean(y_pred == 0), 0.9)

    def test_graph_iterative(self):
        graph = MigrationGraph()
        root = ("app_a", "1")
        graph.add_node(root, None)
        expected = [root]
        for i in range(2, 1000):
            parent = ("app_a", str(i - 1))
            child = ("app_a", str(i))
            graph.add_node(child, None)
            graph.add_dependency(str(i), child, parent)
            expected.append(child)
        leaf = expected[-1]

        with warnings.catch_warnings(record=True) as w:
            warnings.simplefilter('always', RuntimeWarning)
            forwards_plan = graph.forwards_plan(leaf)

        self.assertEqual(len(w), 1)
        self.assertTrue(issubclass(w[-1].category, RuntimeWarning))
        self.assertEqual(str(w[-1].message), RECURSION_DEPTH_WARNING)
        self.assertEqual(expected, forwards_plan)

        with warnings.catch_warnings(record=True) as w:
            warnings.simplefilter('always', RuntimeWarning)
            backwards_plan = graph.backwards_plan(root)

        self.assertEqual(len(w), 1)
        self.assertTrue(issubclass(w[-1].category, RuntimeWarning))
        self.assertEqual(str(w[-1].message), RECURSION_DEPTH_WARNING)
        self.assertEqual(expected[::-1], backwards_plan)

def test_sparse_randomized_pca_inverse():
    """Test that RandomizedPCA is inversible on sparse data"""
    rng = np.random.RandomState(0)
    n, p = 50, 3
    X = rng.randn(n, p)  # spherical data
    X[:, 1] *= 0.00001  # make middle component relatively small
    # no large means because the sparse version of randomized pca does not do
    # centering to avoid breaking the sparsity
    X = csr_matrix(X)

    # same check that we can find the original data from the transformed signal
    # (since the data is almost of rank n_components)
    with warnings.catch_warnings(record=True) as w:
        warnings.simplefilter("always", DeprecationWarning)
        pca = RandomizedPCA(n_components=2, random_state=0).fit(X)
        assert_equal(len(w), 1)
        assert_equal(w[0].category, DeprecationWarning)

    Y = pca.transform(X)

    Y_inverse = pca.inverse_transform(Y)
    assert_almost_equal(X.todense(), Y_inverse, decimal=2)

    # same as above with whitening (approximate reconstruction)
    with warnings.catch_warnings(record=True) as w:
        warnings.simplefilter("always", DeprecationWarning)
        pca = RandomizedPCA(n_components=2, whiten=True, random_state=0).fit(X)
        assert_equal(len(w), 1)
        assert_equal(w[0].category, DeprecationWarning)

    Y = pca.transform(X)
    Y_inverse = pca.inverse_transform(Y)
    relative_max_delta = (np.abs(X.todense() - Y_inverse) / np.abs(X).mean()).max()
    # XXX: this does not seam to work as expected:
    assert_almost_equal(relative_max_delta, 0.91, decimal=2)

    def test_random_pair_match(self) :
        self.assertRaises(ValueError, dedupe.core.randomPairsMatch, 1, 0, 10)
        self.assertRaises(ValueError, dedupe.core.randomPairsMatch, 0, 0, 10)
        self.assertRaises(ValueError, dedupe.core.randomPairsMatch, 0, 1, 10)

        assert len(dedupe.core.randomPairsMatch(100, 100, 100)) == 100
        assert len(dedupe.core.randomPairsMatch(10, 10, 99)) == 99


        with warnings.catch_warnings(record=True) as w:
            warnings.simplefilter("always")
            pairs = dedupe.core.randomPairsMatch(10, 10, 200)
            assert len(w) == 1
            assert str(w[-1].message) == "Requested sample of size 200, only returning 100 possible pairs"

        assert len(pairs) == 100

        with warnings.catch_warnings(record=True) as w:
            warnings.simplefilter("always")
            pairs = dedupe.core.randomPairsMatch(10, 10, 200)
            assert len(w) == 1
            assert str(w[-1].message) == "Requested sample of size 200, only returning 100 possible pairs"


        random.seed(123)
        numpy.random.seed(123)
        pairs = dedupe.core.randomPairsMatch(10, 10, 10)
        assert pairs == set([(7, 3), (3, 3), (2, 9), (6, 0), (2, 0), 
                             (1, 9), (9, 4), (0, 4), (1, 0), (1, 1)])

    def test_random_pair(self) :
        self.assertRaises(ValueError, dedupe.core.randomPairs, 1, 10)
        assert dedupe.core.randomPairs(10, 10).any()
        random.seed(123)
        numpy.random.seed(123)
        random_pairs = dedupe.core.randomPairs(10, 5)
        assert numpy.array_equal(random_pairs, 
                                 numpy.array([[ 0,  3],
                                              [ 3,  8],
                                              [ 4,  9],
                                              [ 5,  9],
                                              [ 2,  3]]))

        with warnings.catch_warnings(record=True) as w:
            warnings.simplefilter("always")
            dedupe.core.randomPairs(10, 10**6)
            assert len(w) == 1
            assert str(w[-1].message) == "Requested sample of size 1000000, only returning 45 possible pairs"

        with warnings.catch_warnings(record=True) as w:
            warnings.simplefilter("always")
            dedupe.core.randomPairs(10**40, 10)
            assert len(w) == 2
            assert str(w[0].message) == "There may be duplicates in the sample"
            assert "Asked to sample pairs from" in str(w[1].message)

        random.seed(123)
        numpy.random.seed(123)
        assert numpy.array_equal(dedupe.core.randomPairs(11**9, 1),
                                 numpy.array([[1228959102, 1840268610]]))

def test_deprecated_score_func():
    # test that old deprecated way of passing a score / loss function is still
    # supported
    X, y = make_classification(n_samples=200, n_features=100, random_state=0)
    clf = LinearSVC(random_state=0)
    cv = GridSearchCV(clf, {'C': [0.1, 1.0]}, scoring="f1")
    cv.fit(X[:180], y[:180])
    y_pred = cv.predict(X[180:])
    C = cv.best_estimator_.C

    clf = LinearSVC(random_state=0)
    cv = GridSearchCV(clf, {'C': [0.1, 1.0]}, score_func=f1_score)
    with warnings.catch_warnings(record=True):
        # catch deprecation warning
        cv.fit(X[:180], y[:180])
    y_pred_func = cv.predict(X[180:])
    C_func = cv.best_estimator_.C

    assert_array_equal(y_pred, y_pred_func)
    assert_equal(C, C_func)

    # test loss where greater is worse
    def f1_loss(y_true_, y_pred_):
        return -f1_score(y_true_, y_pred_)

    clf = LinearSVC(random_state=0)
    cv = GridSearchCV(clf, {'C': [0.1, 1.0]}, loss_func=f1_loss)
    with warnings.catch_warnings(record=True):
        # catch deprecation warning
        cv.fit(X[:180], y[:180])
    y_pred_loss = cv.predict(X[180:])
    C_loss = cv.best_estimator_.C

    assert_array_equal(y_pred, y_pred_loss)
    assert_equal(C, C_loss)