def test_sample_weight_invariance(n_samples=50):
random_state = check_random_state(0)
# binary
random_state = check_random_state(0)
y_true = random_state.randint(0, 2, size=(n_samples, ))
y_pred = random_state.randint(0, 2, size=(n_samples, ))
y_score = random_state.random_sample(size=(n_samples,))
for name in ALL_METRICS:
if (name in METRICS_WITHOUT_SAMPLE_WEIGHT or
name in METRIC_UNDEFINED_BINARY):
continue
metric = ALL_METRICS[name]
if name in THRESHOLDED_METRICS:
yield _named_check(check_sample_weight_invariance, name), name,\
metric, y_true, y_score
else:
yield _named_check(check_sample_weight_invariance, name), name,\
metric, y_true, y_pred
# multiclass
random_state = check_random_state(0)
y_true = random_state.randint(0, 5, size=(n_samples, ))
y_pred = random_state.randint(0, 5, size=(n_samples, ))
y_score = random_state.random_sample(size=(n_samples, 5))
for name in ALL_METRICS:
if (name in METRICS_WITHOUT_SAMPLE_WEIGHT or
name in METRIC_UNDEFINED_BINARY_MULTICLASS):
continue
metric = ALL_METRICS[name]
if name in THRESHOLDED_METRICS:
yield _named_check(check_sample_weight_invariance, name), name,\
metric, y_true, y_score
else:
yield _named_check(check_sample_weight_invariance, name), name,\
metric, y_true, y_pred
# multilabel indicator
_, ya = make_multilabel_classification(n_features=1, n_classes=20,
random_state=0, n_samples=100,
allow_unlabeled=False)
_, yb = make_multilabel_classification(n_features=1, n_classes=20,
random_state=1, n_samples=100,
allow_unlabeled=False)
y_true = np.vstack([ya, yb])
y_pred = np.vstack([ya, ya])
y_score = random_state.randint(1, 4, size=y_true.shape)
for name in (MULTILABELS_METRICS + THRESHOLDED_MULTILABEL_METRICS +
MULTIOUTPUT_METRICS):
if name in METRICS_WITHOUT_SAMPLE_WEIGHT:
continue
metric = ALL_METRICS[name]
if name in THRESHOLDED_METRICS:
yield (_named_check(check_sample_weight_invariance, name), name,
metric, y_true, y_score)
else:
yield (_named_check(check_sample_weight_invariance, name), name,
metric, y_true, y_pred)
def test_normalize_option_multilabel_classification():
# Test in the multilabel case
n_classes = 4
n_samples = 100
# for both random_state 0 and 1, y_true and y_pred has at least one
# unlabelled entry
_, y_true = make_multilabel_classification(n_features=1,
n_classes=n_classes,
random_state=0,
allow_unlabeled=True,
n_samples=n_samples)
_, y_pred = make_multilabel_classification(n_features=1,
n_classes=n_classes,
random_state=1,
allow_unlabeled=True,
n_samples=n_samples)
# To make sure at least one empty label is present
y_true += [0]*n_classes
y_pred += [0]*n_classes
for name in METRICS_WITH_NORMALIZE_OPTION:
metrics = ALL_METRICS[name]
measure = metrics(y_true, y_pred, normalize=True)
assert_greater(measure, 0,
msg="We failed to test correctly the normalize option")
assert_almost_equal(metrics(y_true, y_pred, normalize=False)
/ n_samples, measure,
err_msg="Failed with %s" % name)
def test_multilabel_classification_report():
n_classes = 4
n_samples = 50
_, y_true = make_multilabel_classification(n_features=1,
n_samples=n_samples,
n_classes=n_classes,
random_state=0)
_, y_pred = make_multilabel_classification(n_features=1,
n_samples=n_samples,
n_classes=n_classes,
random_state=1)
expected_report = """\
precision recall f1-score support
0 0.50 0.67 0.57 24
1 0.51 0.74 0.61 27
2 0.29 0.08 0.12 26
3 0.52 0.56 0.54 27
avg / total 0.45 0.51 0.46 104
"""
report = classification_report(y_true, y_pred)
assert_equal(report, expected_report)
def test_normalize_option_multilabel_classification():
# Test in the multilabel case
n_classes = 4
n_samples = 100
_, y_true = make_multilabel_classification(n_features=1, n_classes=n_classes, random_state=0, n_samples=n_samples)
_, y_pred = make_multilabel_classification(n_features=1, n_classes=n_classes, random_state=1, n_samples=n_samples)
# Be sure to have at least one empty label
y_true += ([],)
y_pred += ([],)
n_samples += 1
lb = LabelBinarizer().fit([range(n_classes)])
y_true_binary_indicator = lb.transform(y_true)
y_pred_binary_indicator = lb.transform(y_pred)
for name, metrics in METRICS_WITH_NORMALIZE_OPTION.items():
# List of list of labels
measure = metrics(y_true, y_pred, normalize=True)
assert_greater(measure, 0, msg="We failed to test correctly the normalize option")
assert_almost_equal(
metrics(y_true, y_pred, normalize=False) / n_samples, measure, err_msg="Failed with %s" % name
)
# Indicator matrix format
measure = metrics(y_true_binary_indicator, y_pred_binary_indicator, normalize=True)
assert_greater(measure, 0, msg="We failed to test correctly the normalize option")
assert_almost_equal(
metrics(y_true_binary_indicator, y_pred_binary_indicator, normalize=False) / n_samples,
measure,
err_msg="Failed with %s" % name,
)
def test_multilabel_representation_invariance():
# Generate some data
n_classes = 4
n_samples = 50
_, y1 = make_multilabel_classification(
n_features=1, n_classes=n_classes, random_state=0, n_samples=n_samples, allow_unlabeled=True
)
_, y2 = make_multilabel_classification(
n_features=1, n_classes=n_classes, random_state=1, n_samples=n_samples, allow_unlabeled=True
)
# To make sure at least one empty label is present
y1 += [0] * n_classes
y2 += [0] * n_classes
y1_sparse_indicator = sp.coo_matrix(y1)
y2_sparse_indicator = sp.coo_matrix(y2)
for name in MULTILABELS_METRICS:
metric = ALL_METRICS[name]
# XXX cruel hack to work with partial functions
if isinstance(metric, partial):
metric.__module__ = "tmp"
metric.__name__ = name
measure = metric(y1, y2)
# Check representation invariance
assert_almost_equal(
metric(y1_sparse_indicator, y2_sparse_indicator),
measure,
err_msg="%s failed representation invariance " "between dense and sparse indicator " "formats." % name,
)
def benchmark(metrics=tuple(v for k, v in sorted(METRICS.items())),
formats=tuple(v for k, v in sorted(FORMATS.items())),
samples=1000, classes=4, density=.2,
n_times=5):
"""Times metric calculations for a number of inputs
Parameters
----------
metrics : array-like of callables (1d or 0d)
The metric functions to time.
formats : array-like of callables (1d or 0d)
These may transform a dense indicator matrix into multilabel
representation.
samples : array-like of ints (1d or 0d)
The number of samples to generate as input.
classes : array-like of ints (1d or 0d)
The number of classes in the input.
density : array-like of ints (1d or 0d)
The density of positive labels in the input.
n_times : int
Time calling the metric n_times times.
Returns
-------
array of floats shaped like (metrics, formats, samples, classes, density)
Time in seconds.
"""
metrics = np.atleast_1d(metrics)
samples = np.atleast_1d(samples)
classes = np.atleast_1d(classes)
density = np.atleast_1d(density)
formats = np.atleast_1d(formats)
out = np.zeros((len(metrics), len(formats), len(samples), len(classes),
len(density)), dtype=float)
it = itertools.product(samples, classes, density)
for i, (s, c, d) in enumerate(it):
_, y_true = make_multilabel_classification(n_samples=s, n_features=1,
n_classes=c, n_labels=d * c,
return_indicator=True,
random_state=42)
_, y_pred = make_multilabel_classification(n_samples=s, n_features=1,
n_classes=c, n_labels=d * c,
return_indicator=True,
random_state=84)
for j, f in enumerate(formats):
f_true = f(y_true)
f_pred = f(y_pred)
for k, metric in enumerate(metrics):
t = timeit(partial(metric, f_true, f_pred), number=n_times)
out[k, j].flat[i] = t
return out
def test_ovr_fit_predict_sparse():
for sparse in [sp.csr_matrix, sp.csc_matrix, sp.coo_matrix, sp.dok_matrix, sp.lil_matrix]:
base_clf = MultinomialNB(alpha=1)
X, Y = datasets.make_multilabel_classification(
n_samples=100, n_features=20, n_classes=5, n_labels=3, length=50, allow_unlabeled=True, random_state=0
)
X_train, Y_train = X[:80], Y[:80]
X_test = X[80:]
clf = OneVsRestClassifier(base_clf).fit(X_train, Y_train)
Y_pred = clf.predict(X_test)
clf_sprs = OneVsRestClassifier(base_clf).fit(X_train, sparse(Y_train))
Y_pred_sprs = clf_sprs.predict(X_test)
assert_true(clf.multilabel_)
assert_true(sp.issparse(Y_pred_sprs))
assert_array_equal(Y_pred_sprs.toarray(), Y_pred)
# Test predict_proba
Y_proba = clf_sprs.predict_proba(X_test)
# predict assigns a label if the probability that the
# sample has the label is greater than 0.5.
pred = Y_proba > 0.5
assert_array_equal(pred, Y_pred_sprs.toarray())
# Test decision_function
clf_sprs = OneVsRestClassifier(svm.SVC()).fit(X_train, sparse(Y_train))
dec_pred = (clf_sprs.decision_function(X_test) > 0).astype(int)
assert_array_equal(dec_pred, clf_sprs.predict(X_test).toarray())
def test_ovr_multilabel_predict_proba():
base_clf = MultinomialNB(alpha=1)
for au in (False, True):
X, Y = datasets.make_multilabel_classification(n_samples=100,
n_features=20,
n_classes=5,
n_labels=3,
length=50,
allow_unlabeled=au,
return_indicator=True,
random_state=0)
X_train, Y_train = X[:80], Y[:80]
X_test, Y_test = X[80:], Y[80:]
clf = OneVsRestClassifier(base_clf).fit(X_train, Y_train)
# decision function only estimator. Fails in current implementation.
decision_only = OneVsRestClassifier(svm.SVR()).fit(X_train, Y_train)
assert_raises(AttributeError, decision_only.predict_proba, X_test)
# Estimator with predict_proba disabled, depending on parameters.
decision_only = OneVsRestClassifier(svm.SVC(probability=False))
decision_only.fit(X_train, Y_train)
assert_raises(AttributeError, decision_only.predict_proba, X_test)
Y_pred = clf.predict(X_test)
Y_proba = clf.predict_proba(X_test)
# predict assigns a label if the probability that the
# sample has the label is greater than 0.5.
pred = Y_proba > .5
assert_array_equal(pred, Y_pred)
def check_alternative_lrap_implementation(lrap_score, n_classes=5,
n_samples=20, random_state=0):
_, y_true = make_multilabel_classification(n_features=1,
allow_unlabeled=False,
random_state=random_state,
n_classes=n_classes,
n_samples=n_samples)
# Score with ties
y_score = sparse_random_matrix(n_components=y_true.shape[0],
n_features=y_true.shape[1],
random_state=random_state)
if hasattr(y_score, "toarray"):
y_score = y_score.toarray()
score_lrap = label_ranking_average_precision_score(y_true, y_score)
score_my_lrap = _my_lrap(y_true, y_score)
assert_almost_equal(score_lrap, score_my_lrap)
# Uniform score
random_state = check_random_state(random_state)
y_score = random_state.uniform(size=(n_samples, n_classes))
score_lrap = label_ranking_average_precision_score(y_true, y_score)
score_my_lrap = _my_lrap(y_true, y_score)
assert_almost_equal(score_lrap, score_my_lrap)
def get_multilabel(self):
return make_multilabel_classification(n_samples=100,
n_features=10,
n_classes=5,
n_labels=5,
return_indicator=True,
random_state=1)
def testMultiClassification(self):
"""TODO(ilblackdragon): Implement multi-output classification.
"""
random.seed(42)
n_classes = 5
X, y = datasets.make_multilabel_classification(n_classes=n_classes,
random_state=42)
def get_codes(self):
X, Y = make_multilabel_classification(n_samples=15, n_labels=8, n_classes=8, random_state=0)
self.classifier_labels = Y
self.classifier_error_codes = LabelBinarizer().fit_transform(Y)
print self.classifier_labels
print self.classifier_error_codes
f = open('ecoc_classifiers', 'w')
for row in self.classifier_labels:
str_op = '['
for label in row:
str_op += str(label) + ','
str_op += ']'
f.write(str_op)
f.write('\n')
for row in self.classifier_error_codes:
str_op = '['
for label in row:
str_op += str(label) + ','
str_op += ']'
f.write(str_op)
f.flush()
return
def test_sparse_input():
X, y = datasets.make_multilabel_classification(random_state=0,
n_samples=50)
for name, sparse_matrix in product(FOREST_ESTIMATORS,
(csr_matrix, csc_matrix, coo_matrix)):
yield check_sparse_input, name, X, sparse_matrix(X), y
def test_random_hasher_sparse_data():
X, y = datasets.make_multilabel_classification(return_indicator=True,
random_state=0)
hasher = RandomTreesEmbedding(n_estimators=30, random_state=1)
X_transformed = hasher.fit_transform(X)
X_transformed_sparse = hasher.fit_transform(csc_matrix(X))
assert_array_equal(X_transformed_sparse.toarray(), X_transformed.toarray())
def test_output_transformer():
X, y = datasets.make_multilabel_classification(return_indicator=True)
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=1)
# Check that random_state are different
transformer = GaussianRandomProjection(n_components=5, random_state=None)
for name, ForestEstimator in FOREST_ESTIMATORS.items():
est = ForestEstimator(random_state=5, output_transformer=transformer)
est.fit(X_train, y_train)
y_pred = est.predict(X_test)
assert_equal(y_pred.shape, y_test.shape)
random_state = [sub.output_transformer_.random_state
for sub in est.estimators_]
assert_equal(len(set(random_state)), est.n_estimators)
# Check that random_state are equals
transformer = FixedStateTransformer(GaussianRandomProjection(
n_components=5), random_seed=0)
for name, ForestEstimator in FOREST_ESTIMATORS.items():
est = ForestEstimator(random_state=5, output_transformer=transformer)
est.fit(X_train, y_train)
y_pred = est.predict(X_test)
assert_equal(y_pred.shape, y_test.shape)
random_state = [sub.output_transformer_.random_state
for sub in est.estimators_]
assert_equal(len(set(random_state)), 1)
assert_equal(random_state[0], 0)
def test_grid_search_with_multioutput_data():
# Test search with multi-output estimator
X, y = make_multilabel_classification(random_state=0)
est_parameters = {"max_depth": [1, 2, 3, 4]}
cv = KFold(y.shape[0], random_state=0)
estimators = [DecisionTreeRegressor(random_state=0), DecisionTreeClassifier(random_state=0)]
# Test with grid search cv
for est in estimators:
grid_search = GridSearchCV(est, est_parameters, cv=cv)
grid_search.fit(X, y)
for parameters, _, cv_validation_scores in grid_search.grid_scores_:
est.set_params(**parameters)
for i, (train, test) in enumerate(cv):
est.fit(X[train], y[train])
correct_score = est.score(X[test], y[test])
assert_almost_equal(correct_score, cv_validation_scores[i])
# Test with a randomized search
for est in estimators:
random_search = RandomizedSearchCV(est, est_parameters, cv=cv, n_iter=3)
random_search.fit(X, y)
for parameters, _, cv_validation_scores in random_search.grid_scores_:
est.set_params(**parameters)
for i, (train, test) in enumerate(cv):
est.fit(X[train], y[train])
correct_score = est.score(X[test], y[test])
assert_almost_equal(correct_score, cv_validation_scores[i])
def test_multilabel_classification():
"""Test that multi-label classification works as expected."""
# test fit method
X, y = make_multilabel_classification(n_samples=50, random_state=0,
return_indicator=True)
elm = ELMClassifier(weight_scale=100)
elm.fit(X, y)
assert_greater(elm.score(X, y), 0.95)
def assertClassifierWorksWithSparsity(self, classifier, sparsity_indicator = 'sparse'):
feed_sparse = sparsity_indicator == 'sparse'
X, y = make_multilabel_classification(sparse = feed_sparse, return_indicator = sparsity_indicator)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=42)
classifier.fit(X_train, y_train)
result = classifier.predict(X_test)
self.assertEqual(result.shape, y_test.shape)
def test_ovr_multilabel_decision_function():
X, Y = datasets.make_multilabel_classification(
n_samples=100, n_features=20, n_classes=5, n_labels=3, length=50, allow_unlabeled=True, random_state=0
)
X_train, Y_train = X[:80], Y[:80]
X_test = X[80:]
clf = OneVsRestClassifier(svm.SVC()).fit(X_train, Y_train)
assert_array_equal((clf.decision_function(X_test) > 0).astype(int), clf.predict(X_test))
def test_sparse_input(EstimatorClass, sparse_matrix):
y, X = datasets.make_multilabel_classification(random_state=0,
n_samples=50,
n_features=1,
n_classes=20)
y = y[:, 0]
check_sparse_input(EstimatorClass, X, sparse_matrix(X), y)
def test_make_multilabel_classification_return_indicator():
for allow_unlabeled, min_length in zip((True, False), (0, 1)):
X, Y = make_multilabel_classification(n_samples=25, n_features=20,
n_classes=3, random_state=0,
return_indicator=True,
allow_unlabeled=allow_unlabeled)
assert_equal(X.shape, (25, 20), "X shape mismatch")
assert_equal(Y.shape, (25, 3), "Y shape mismatch")
assert_true(np.all(np.sum(Y, axis=0) > min_length))
def test_multilabel_sample_weight_invariance(name):
# multilabel indicator
random_state = check_random_state(0)
_, ya = make_multilabel_classification(n_features=1, n_classes=20,
random_state=0, n_samples=100,
allow_unlabeled=False)
_, yb = make_multilabel_classification(n_features=1, n_classes=20,
random_state=1, n_samples=100,
allow_unlabeled=False)
y_true = np.vstack([ya, yb])
y_pred = np.vstack([ya, ya])
y_score = random_state.randint(1, 4, size=y_true.shape)
metric = ALL_METRICS[name]
if name in THRESHOLDED_METRICS:
check_sample_weight_invariance(name, metric, y_true, y_score)
else:
check_sample_weight_invariance(name, metric, y_true, y_pred)
def test_make_multilabel_classification_return_indicator_sparse():
for allow_unlabeled, min_length in zip((True, False), (0, 1)):
X, Y = make_multilabel_classification(n_samples=25, n_features=20,
n_classes=3, random_state=0,
return_indicator='sparse',
allow_unlabeled=allow_unlabeled)
assert_equal(X.shape, (25, 20), "X shape mismatch")
assert_equal(Y.shape, (25, 3), "Y shape mismatch")
assert_true(sp.issparse(Y))
def test_scorer_sample_weight():
# Test that scorers support sample_weight or raise sensible errors
# Unlike the metrics invariance test, in the scorer case it's harder
# to ensure that, on the classifier output, weighted and unweighted
# scores really should be unequal.
X, y = make_classification(random_state=0)
_, y_ml = make_multilabel_classification(n_samples=X.shape[0], random_state=0)
split = train_test_split(X, y, y_ml, random_state=0)
X_train, X_test, y_train, y_test, y_ml_train, y_ml_test = split
sample_weight = np.ones_like(y_test)
sample_weight[:10] = 0
# get sensible estimators for each metric
sensible_regr = DummyRegressor(strategy="median")
sensible_regr.fit(X_train, y_train)
sensible_clf = DecisionTreeClassifier(random_state=0)
sensible_clf.fit(X_train, y_train)
sensible_ml_clf = DecisionTreeClassifier(random_state=0)
sensible_ml_clf.fit(X_train, y_ml_train)
estimator = dict(
[(name, sensible_regr) for name in REGRESSION_SCORERS]
+ [(name, sensible_clf) for name in CLF_SCORERS]
+ [(name, sensible_ml_clf) for name in MULTILABEL_ONLY_SCORERS]
)
for name, scorer in SCORERS.items():
if name in MULTILABEL_ONLY_SCORERS:
target = y_ml_test
else:
target = y_test
try:
weighted = scorer(estimator[name], X_test, target, sample_weight=sample_weight)
ignored = scorer(estimator[name], X_test[10:], target[10:])
unweighted = scorer(estimator[name], X_test, target)
assert_not_equal(
weighted,
unweighted,
msg="scorer {0} behaves identically when "
"called with sample weights: {1} vs "
"{2}".format(name, weighted, unweighted),
)
assert_almost_equal(
weighted,
ignored,
err_msg="scorer {0} behaves differently when "
"ignoring samples and setting sample_weight to"
" 0: {1} vs {2}".format(name, weighted, ignored),
)
except TypeError as e:
assert_true(
"sample_weight" in str(e),
"scorer {0} raises unhelpful exception when called " "with sample weights: {1}".format(name, str(e)),
)
def test_class_type(self):
"""
Test class must be either binary or multiclass type
"""
X, y = make_multilabel_classification()
model = RandomForestClassifier()
model.fit(X, y)
with self.assertRaises(YellowbrickValueError):
visualizer = ClassPredictionError(model)
visualizer.score(X, y)
def test_output_transformer():
X, y = datasets.make_multilabel_classification(return_indicator=True)
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=1)
transformer = GaussianRandomProjection(n_components=10)
for name, TreeEstimator in ALL_TREES.items():
est = TreeEstimator(random_state=0, output_transformer=transformer)
est.fit(X_train, y_train)
y_pred = est.predict(X_test)
assert_equal(y_pred.shape, y_test.shape)
def test():
X,Y_list = make_multilabel_classification()
Y = LabelBinarizer().fit_transform(Y_list)
Y[Y==0] = -1
clf = OneVsRestClassifier(LinearSVC())
#clf = MultilabelLR(L0=1, λ1=0.1, λ2=0.1, γ=0.1, μ=0.1)
clf.fit(X,Y)
Y_hat = clf.predict(X)
print(roc_auc_score(Y.flat, Y_hat.flat))
def test_make_multilabel_classification():
for allow_unlabeled, min_length in zip((True, False), (0, 1)):
X, Y = make_multilabel_classification(n_samples=100, n_features=20,
n_classes=3, random_state=0,
allow_unlabeled=allow_unlabeled)
assert_equal(X.shape, (100, 20), "X shape mismatch")
if not allow_unlabeled:
assert_equal(max([max(y) for y in Y]), 2)
assert_equal(min([len(y) for y in Y]), min_length)
assert_true(max([len(y) for y in Y]) <= 3)
def setup_module():
# Create some memory mapped data
global X_mm, y_mm, y_ml_mm, TEMP_FOLDER, ESTIMATORS
TEMP_FOLDER = tempfile.mkdtemp(prefix='sklearn_test_score_objects_')
X, y = make_classification(n_samples=30, n_features=5, random_state=0)
_, y_ml = make_multilabel_classification(n_samples=X.shape[0],
random_state=0)
filename = os.path.join(TEMP_FOLDER, 'test_data.pkl')
joblib.dump((X, y, y_ml), filename)
X_mm, y_mm, y_ml_mm = joblib.load(filename, mmap_mode='r')
ESTIMATORS = _make_estimators(X_mm, y_mm, y_ml_mm)
def test_make_multilabel_classification_return_indicator():
for allow_unlabeled, min_length in zip((True, False), (0, 1)):
X, Y = make_multilabel_classification(n_samples=25, n_features=20,
n_classes=3, random_state=0,
allow_unlabeled=allow_unlabeled)
assert_equal(X.shape, (25, 20), "X shape mismatch")
assert_equal(Y.shape, (25, 3), "Y shape mismatch")
assert_true(np.all(np.sum(Y, axis=0) > min_length))
# Also test return_distributions and return_indicator with True
X2, Y2, p_c, p_w_c = make_multilabel_classification(
n_samples=25, n_features=20, n_classes=3, random_state=0,
allow_unlabeled=allow_unlabeled, return_distributions=True)
assert_array_equal(X, X2)
assert_array_equal(Y, Y2)
assert_equal(p_c.shape, (3,))
assert_almost_equal(p_c.sum(), 1)
assert_equal(p_w_c.shape, (20, 3))
assert_almost_equal(p_w_c.sum(axis=0), [1] * 3)
def test_ovr_fit_predict_sparse():
for sparse in [
sp.csr_matrix, sp.csc_matrix, sp.coo_matrix, sp.dok_matrix,
sp.lil_matrix
]:
base_clf = MultinomialNB(alpha=1)
X, Y = datasets.make_multilabel_classification(n_samples=100,
n_features=20,
n_classes=5,
n_labels=3,
length=50,
allow_unlabeled=True,
random_state=0)
X_train, Y_train = X[:80], Y[:80]
X_test = X[80:]
clf = OneVsRestClassifier(base_clf).fit(X_train, Y_train)
Y_pred = clf.predict(X_test)
clf_sprs = OneVsRestClassifier(base_clf).fit(X_train, sparse(Y_train))
Y_pred_sprs = clf_sprs.predict(X_test)
assert_true(clf.multilabel_)
assert_true(sp.issparse(Y_pred_sprs))
assert_array_equal(Y_pred_sprs.toarray(), Y_pred)
# Test predict_proba
Y_proba = clf_sprs.predict_proba(X_test)
# predict assigns a label if the probability that the
# sample has the label is greater than 0.5.
pred = Y_proba > .5
assert_array_equal(pred, Y_pred_sprs.toarray())
# Test decision_function
clf_sprs = OneVsRestClassifier(svm.SVC()).fit(X_train, sparse(Y_train))
dec_pred = (clf_sprs.decision_function(X_test) > 0).astype(int)
assert_array_equal(dec_pred, clf_sprs.predict(X_test).toarray())
def test_ovr_multilabel_predict_proba():
base_clf = MultinomialNB(alpha=1)
for au in (False, True):
X, Y = datasets.make_multilabel_classification(n_samples=100,
n_features=20,
n_classes=5,
n_labels=3,
length=50,
allow_unlabeled=au,
random_state=0)
X_train, Y_train = X[:80], Y[:80]
X_test = X[80:]
clf = OneVsRestClassifier(base_clf).fit(X_train, Y_train)
# Decision function only estimator.
decision_only = OneVsRestClassifier(svm.SVR()).fit(X_train, Y_train)
assert not hasattr(decision_only, 'predict_proba')
# Estimator with predict_proba disabled, depending on parameters.
decision_only = OneVsRestClassifier(svm.SVC(probability=False))
assert not hasattr(decision_only, 'predict_proba')
decision_only.fit(X_train, Y_train)
assert not hasattr(decision_only, 'predict_proba')
assert hasattr(decision_only, 'decision_function')
# Estimator which can get predict_proba enabled after fitting
gs = GridSearchCV(svm.SVC(probability=False),
param_grid={'probability': [True]})
proba_after_fit = OneVsRestClassifier(gs)
assert not hasattr(proba_after_fit, 'predict_proba')
proba_after_fit.fit(X_train, Y_train)
assert hasattr(proba_after_fit, 'predict_proba')
Y_pred = clf.predict(X_test)
Y_proba = clf.predict_proba(X_test)
# predict assigns a label if the probability that the
# sample has the label is greater than 0.5.
pred = Y_proba > .5
assert_array_equal(pred, Y_pred)
def test_grid_search_with_multioutput_data():
""" Test search with multi-output estimator"""
X, y = make_multilabel_classification(return_indicator=True,
random_state=0)
est_parameters = {"max_depth": [1, 2, 3, 4]}
cv = KFold(y.shape[0], random_state=0)
estimators = [
DecisionTreeRegressor(random_state=0),
DecisionTreeClassifier(random_state=0)
]
# Test with grid search cv
for est in estimators:
grid_search = GridSearchCV(est, est_parameters, cv=cv)
grid_search.fit(X, y)
for parameters, _, cv_validation_scores in grid_search.grid_scores_:
est.set_params(**parameters)
for i, (train, test) in enumerate(cv):
est.fit(X[train], y[train])
correct_score = est.score(X[test], y[test])
assert_almost_equal(correct_score, cv_validation_scores[i])
# Test with a randomized search
for est in estimators:
random_search = RandomizedSearchCV(est,
est_parameters,
cv=cv,
n_iter=3)
random_search.fit(X, y)
for parameters, _, cv_validation_scores in random_search.grid_scores_:
est.set_params(**parameters)
for i, (train, test) in enumerate(cv):
est.fit(X[train], y[train])
correct_score = est.score(X[test], y[test])
assert_almost_equal(correct_score, cv_validation_scores[i])
def test_multioutput():
X, y = make_multilabel_classification(n_samples=100, n_labels=1,
n_classes=5, random_state=0,
return_indicator=True)
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)
est = LazyBaggingClassifier(random_state=0, n_estimators=10,
bootstrap=False)
est.fit(X_train, y_train)
assert_almost_equal(est.score(X_train, y_train), 1.)
y_proba = est.predict_proba(X_test)
y_log_proba = est.predict_log_proba(X_test)
for p, log_p in zip(y_proba, y_log_proba):
assert_array_almost_equal(p, np.exp(log_p))
est = LazyBaggingRegressor(random_state=0, n_estimators=10,
bootstrap=False)
est.fit(X_train, y_train)
assert_almost_equal(est.score(X_train, y_train), 1.)
def multilLabel():
# 多标签多分类原始标签
y = [[2, 3, 4], [2], [0, 1, 3], [0, 1, 2, 3, 4], [0, 1, 2]]
# 对标签进行预处理
mb = MultiLabelBinarizer()
# y_mb变成N*K的矩阵(N:样本数,K:类别数)
y_mb = mb.fit_transform(y)
# 多类别学习,标签形如[0,0,1,1,2,2]
data = load_iris()
X, y = data.data, data.target
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.2,
random_state=0)
binary_model = SVC(kernel="linear", random_state=1)
# one-vs-all形式(既可以多类别问题也可以多标签多分类问题，fir(X,y)中y.shape=[samples] or [samples,classes])
multi_model = OneVsRestClassifier(binary_model).fit(X_train, y_train)
# one-vs-one形式(只能用于多类别问题，fit(X,y)函数要求y.shape=[samples])
#multi_model = OneVsOneClassifier(binary_model).fit(X_train,y_train)
y_pred = multi_model.predict(X_test)
print("True Labels: ", y_test)
print("Predict Labels:", y_pred)
print("Accuracy: ", accuracy_score(y_test, y_pred))
# 多标签多分类
ml_X, ml_y = make_multilabel_classification()
print("多标签多分类训练标签:\n", ml_y[:5])
ml_X_train, ml_X_test, ml_y_train, ml_y_test = train_test_split(
ml_X, ml_y, test_size=0.1)
# one-vs-all
clf = OneVsRestClassifier(SVC(kernel="linear"))
clf.fit(ml_X_train, ml_y_train)
pred_y = clf.predict(ml_X_test)
print("True Labels: \n", ml_y_test)
print("Predict Labels:\n", pred_y)
print("Hamming_loss: ", hamming_loss(ml_y_test, pred_y))
print("Accuracy: ", accuracy_score(ml_y_test, pred_y))
def test_evenly_distributes_unlabelled(self):
cv = IterativeStratifiedKFold(n_splits=5,
shuffle=False,
random_state=0)
X, y = make_multilabel_classification(100,
20,
n_labels=5,
random_state=0,
allow_unlabeled=False)
y[[0, 1, 2, 3, 4], :] = 0
# Make label 0 have only 3 positive instances.
folds = list(cv.split(X, y))
for train_idx, valid_idx in folds:
unlabelled_in_train = np.where(
np.sum(y[train_idx, :], axis=1) == 0)[0].shape[0]
unlabelled_in_valid = np.where(
np.sum(y[valid_idx, :], axis=1) == 0)[0].shape[0]
self.assertEqual(unlabelled_in_train, 4)
self.assertEqual(unlabelled_in_valid, 1)
def generate_classification(self,
num_classes,
num_features,
num_samples,
test_split=0.1,
seed=0):
#X, Y = make_classification(n_samples=800, n_features=num_feats, n_classes=num_classes, n_informative=4)
X, y = make_multilabel_classification(n_samples=num_samples,
n_features=num_features,
n_classes=num_classes,
n_labels=0.01,
length=50,
allow_unlabeled=False,
sparse=False,
return_indicator='dense',
return_distributions=False,
random_state=seed)
Y = np.argmax(y, axis=1)
self.categorical_features = [False] * num_features
self.problem_type = ProblemType.FeatureClassification
self.X, self.Y = X, Y
self._split_data(test_split, seed)
def dataset(request):
X, y = make_multilabel_classification(
n_samples=int(request.param['n_samples'] * 1.2),
n_features=request.param['n_features'],
n_classes=request.param['n_classes'],
n_labels=request.param['n_classes'],
length=request.param['n_targets'])
new_x = []
new_y = []
for i in range(y.shape[0]):
a = np.argwhere(y[i] == 1)[:, 0]
if len(a) >= request.param['n_targets']:
new_x.append(i)
np.random.shuffle(a)
a = a[:request.param['n_targets']]
new_y.append(a)
if len(new_x) >= request.param['n_samples']:
break
X = X[new_x]
y = np.array(new_y)
return train_test_split(X, y, test_size=0.33)
def test_check_classifiers_multilabel_output_format_predict():
n_samples, test_size, n_outputs = 100, 25, 5
_, y = make_multilabel_classification(
n_samples=n_samples,
n_features=2,
n_classes=n_outputs,
n_labels=3,
length=50,
allow_unlabeled=True,
random_state=0,
)
y_test = y[-test_size:]
class MultiLabelClassifierPredict(_BaseMultiLabelClassifierMock):
def predict(self, X):
return self.response_output
# 1. inconsistent array type
clf = MultiLabelClassifierPredict(response_output=y_test.tolist())
err_msg = (r"MultiLabelClassifierPredict.predict is expected to output a "
r"NumPy array. Got <class 'list'> instead.")
with raises(AssertionError, match=err_msg):
check_classifiers_multilabel_output_format_predict(
clf.__class__.__name__, clf)
# 2. inconsistent shape
clf = MultiLabelClassifierPredict(response_output=y_test[:, :-1])
err_msg = (r"MultiLabelClassifierPredict.predict outputs a NumPy array of "
r"shape \(25, 4\) instead of \(25, 5\).")
with raises(AssertionError, match=err_msg):
check_classifiers_multilabel_output_format_predict(
clf.__class__.__name__, clf)
# 3. inconsistent dtype
clf = MultiLabelClassifierPredict(
response_output=y_test.astype(np.float64))
err_msg = (r"MultiLabelClassifierPredict.predict does not output the same "
r"dtype than the targets.")
with raises(AssertionError, match=err_msg):
check_classifiers_multilabel_output_format_predict(
clf.__class__.__name__, clf)
def test_predict_proba_multilabel():
# Test that predict_proba works as expected for multilabel.
# Multilabel should not use softmax which makes probabilities sum to 1
X, Y = make_multilabel_classification(n_samples=50,
random_state=0,
return_indicator=True)
n_samples, n_classes = Y.shape
clf = MLPClassifier(solver='lbfgs', hidden_layer_sizes=30, random_state=0)
clf.fit(X, Y)
y_proba = clf.predict_proba(X)
assert_equal(y_proba.shape, (n_samples, n_classes))
assert_array_equal(y_proba > 0.5, Y)
y_log_proba = clf.predict_log_proba(X)
proba_max = y_proba.argmax(axis=1)
proba_log_max = y_log_proba.argmax(axis=1)
assert_greater((y_proba.sum(1) - 1).dot(y_proba.sum(1) - 1), 1e-10)
assert_array_equal(proba_max, proba_log_max)
assert_array_equal(y_log_proba, np.log(y_proba))
def test_ovr_multilabel_dataset():
base_clf = MultinomialNB(alpha=1)
for au, prec, recall in zip((True, False), (0.51, 0.66), (0.51, 0.80)):
X, Y = datasets.make_multilabel_classification(n_samples=100,
n_features=20,
n_classes=5,
n_labels=2,
length=50,
allow_unlabeled=au,
random_state=0)
X_train, Y_train = X[:80], Y[:80]
X_test, Y_test = X[80:], Y[80:]
clf = OneVsRestClassifier(base_clf).fit(X_train, Y_train)
Y_pred = clf.predict(X_test)
assert_true(clf.multilabel_)
assert_almost_equal(precision_score(Y_test, Y_pred, average="micro"),
prec,
decimal=2)
assert_almost_equal(recall_score(Y_test, Y_pred, average="micro"),
recall,
decimal=2)
def test_sparse_input(EstimatorClass, sparse_matrix):
y, X = datasets.make_multilabel_classification(
random_state=0, n_samples=50, n_features=1, n_classes=20
)
y = y[:, 0]
X_sparse = sparse_matrix(X)
dense = EstimatorClass(
n_estimators=10, random_state=0, max_depth=2, min_impurity_decrease=1e-7
).fit(X, y)
sparse = EstimatorClass(
n_estimators=10, random_state=0, max_depth=2, min_impurity_decrease=1e-7
).fit(X_sparse, y)
assert_array_almost_equal(sparse.apply(X), dense.apply(X))
assert_array_almost_equal(sparse.predict(X), dense.predict(X))
assert_array_almost_equal(sparse.feature_importances_, dense.feature_importances_)
assert_array_almost_equal(sparse.predict(X_sparse), dense.predict(X))
assert_array_almost_equal(dense.predict(X_sparse), sparse.predict(X))
if issubclass(EstimatorClass, GradientBoostingClassifier):
assert_array_almost_equal(sparse.predict_proba(X), dense.predict_proba(X))
assert_array_almost_equal(
sparse.predict_log_proba(X), dense.predict_log_proba(X)
)
assert_array_almost_equal(
sparse.decision_function(X_sparse), sparse.decision_function(X)
)
assert_array_almost_equal(
dense.decision_function(X_sparse), sparse.decision_function(X)
)
for res_sparse, res in zip(
sparse.staged_decision_function(X_sparse),
sparse.staged_decision_function(X),
):
assert_array_almost_equal(res_sparse, res)
def test_actually_works_on_proper_params(self):
X, y = make_multilabel_classification(sparse=True,
return_indicator='sparse')
assert sp.issparse(y)
for allow_overlap in [True, False]:
for weighted in [True, False]:
for include_self_edges in [True, False]:
for use_degree_corr in [True, False, None]:
for model_selection_criterium in [
'mean_field', 'bethe'
]:
for verbose in [True, False]:
clusterer = GraphToolCooccurenceClusterer(
weighted=weighted,
allow_overlap=allow_overlap,
include_self_edges=include_self_edges,
n_iters=2,
n_init_iters=2,
use_degree_corr=use_degree_corr,
model_selection_criterium=
model_selection_criterium,
verbose=verbose)
self.assertEqual(clusterer.allow_overlap,
allow_overlap)
self.assertEqual(clusterer.is_weighted,
weighted)
self.assertEqual(clusterer.include_self_edges,
include_self_edges)
self.assertEqual(clusterer.n_iters, 2)
self.assertEqual(clusterer.n_init_iters, 2)
self.assertEqual(
clusterer.model_selection_criterium,
model_selection_criterium)
self.assertEqual(clusterer.verbose, verbose)
partition = clusterer.fit_predict(X, y)
self.assertIsInstance(partition, np.ndarray)
def test_evenly_distributes_label_with_multilabel(self):
X, y = make_multilabel_classification(100,
20,
n_labels=18,
random_state=0,
allow_unlabeled=False,
n_classes=18)
d_idx = 0
c_idx = 1
y[:, d_idx] = 0
y[:, c_idx] = 0
y[[0, 10, 20, 12, 4], d_idx] = 1
y[[1, 11, 21, 12, 4], c_idx] = 1
# With Shuffle
iskf = IterativeStratifiedKFold(n_splits=3,
shuffle=True,
random_state=42)
cv = list(iskf.split(X, y))
for train, valid in cv:
self.assertIn(y[train].sum(axis=0)[d_idx], (4, 3))
self.assertIn(y[valid].sum(axis=0)[d_idx], (1, 2))
for train, valid in cv:
self.assertIn(y[train].sum(axis=0)[c_idx], (4, 3))
self.assertIn(y[valid].sum(axis=0)[c_idx], (1, 2))
# With Shuffle
iskf = IterativeStratifiedKFold(n_splits=3,
shuffle=False,
random_state=42)
cv = list(iskf.split(X, y))
for train, valid in cv:
self.assertIn(y[train].sum(axis=0)[d_idx], (4, 3))
self.assertIn(y[valid].sum(axis=0)[d_idx], (1, 2))
for train, valid in cv:
self.assertIn(y[train].sum(axis=0)[c_idx], (4, 3))
self.assertIn(y[valid].sum(axis=0)[c_idx], (1, 2))
def __configure(self):
""" __configure
Uses the make_multilabel_classification function from scikit-learn
to generate a multilabel classification problem. This problem will
be kept in memory and provided as demanded.
"""
self.X, self.y = make_multilabel_classification(
n_samples=self.n_samples,
n_features=self.n_features,
n_classes=self.n_targets,
n_labels=self.n_labels,
random_state=self.random_state)
self.target_names = ["target_" + str(i) for i in range(self.n_targets)]
self.feature_names = [
"att_num_" + str(i) for i in range(self.n_num_features)
]
self.target_values = np.unique(
self.y).tolist() if self.n_targets == 1 else [
np.unique(self.y[:, i]).tolist() for i in range(self.n_targets)
]
def setUp(self):
self.binary_problem_instance = make_classification(n_classes=2, n_samples=100, n_features=10, \
n_informative=8, n_redundant=0,random_state=0, shuffle=False)
self.Xb = self.binary_problem_instance[0]
self.yb = self.binary_problem_instance[1]
self.Xb, self.yb = helpers.binary_to_regression(self.Xb, self.yb)
self.multiClass_problem_instance = make_classification(n_classes=4, n_samples=100, n_features=10, \
n_informative=8, n_redundant=0,random_state=0, shuffle=False)
self.Xmc = self.multiClass_problem_instance[0]
self.ymc = self.multiClass_problem_instance[1]
self.Xmc, self.ymc = helpers.multiClass_to_regression(
self.Xmc, self.ymc, 4)
self.multiLabel_problem_instance = make_multilabel_classification(n_classes=5, \
n_labels=2, n_samples=100, n_features=10)
self.Xml = self.multiLabel_problem_instance[0]
self.yml = self.multiLabel_problem_instance[1]
self.Xml, self.yml = helpers.multiLabel_to_regression(
self.Xml, self.yml, 5)
def test_sparse():
""" Validate running LinearExplainer on scipy sparse data
"""
import sklearn.linear_model
from sklearn.datasets import make_multilabel_classification
from scipy.special import expit
np.random.seed(0)
n_features = 20
X, y = make_multilabel_classification(n_samples=100,
sparse=True,
n_features=n_features,
n_classes=1,
n_labels=2)
# train linear model
model = sklearn.linear_model.LogisticRegression()
model.fit(X, y)
# explain the model's predictions using SHAP values
explainer = shap.LinearExplainer(model, X)
shap_values = explainer.shap_values(X)
assert np.max(np.abs(expit(explainer.expected_value + shap_values.sum(1)) - model.predict_proba(X)[:, 1])) < 1e-6
def get_data():
x,y = make_multilabel_classification(n_samples=20,n_features=2,\
n_labels=1,n_classes=1,random_state=2)
# #创建表格
# wb = workbook.Workbook()
# #表示动作句柄
# wa = wb.active
# for i in range(len(x)):
# # print(list(x[i])+list(y[i]))
# wa.append(list(x[i])+list(y[i]))
# wb.save('data.xlsx')
# # read_excel_xlsx('data.xlsx')
# x：特征值，y：类别
# 根据类别分个类
# 类别1的下标
index1 = np.array([index for (index,value) in enumerate(y) if value == 0]) # print(index1)
#类别2的下标
index2 = np.array([index for (index,value) in enumerate(y) if value == 1])
c1 = x[index1]
c2 = x[index2]
return x,np.array([c1,c2])
def test_multilabel_classification():
# Test that multi-label classification works as expected.
# test fit method
X, y = make_multilabel_classification(n_samples=50, random_state=0,
return_indicator=True)
mlp = MLPClassifier(solver='lbfgs', hidden_layer_sizes=50, alpha=1e-5,
max_iter=150, random_state=0, activation='logistic',
learning_rate_init=0.2)
mlp.fit(X, y)
assert mlp.score(X, y) > 0.97
# test partial fit method
mlp = MLPClassifier(solver='sgd', hidden_layer_sizes=50, max_iter=150,
random_state=0, activation='logistic', alpha=1e-5,
learning_rate_init=0.2)
for i in range(100):
mlp.partial_fit(X, y, classes=[0, 1, 2, 3, 4])
assert mlp.score(X, y) > 0.9
# Make sure early stopping still work now that spliting is stratified by
# default (it is disabled for multilabel classification)
mlp = MLPClassifier(early_stopping=True)
mlp.fit(X, y).predict(X)
def generate_classification(self, num_classes, num_features, num_samples, test_split=0.1, seed=0):
"""Generate a classification task
Arguments:
num_classes {int} -- Number of classes
num_features {int} -- Number of features
num_samples {int} -- Number of samples
Keyword Arguments:
test_split {float} -- Size of test split (default: {0.1})
seed {int} -- A random seed (default: {0})
"""
#X, Y = make_classification(n_samples=800, n_features=num_feats, n_classes=num_classes, n_informative=4)
X, y = make_multilabel_classification(
n_samples=num_samples, n_features=num_features, n_classes=num_classes, n_labels=0.01,
length=50, allow_unlabeled=False, sparse=False, return_indicator='dense',
return_distributions=False, random_state=seed
)
Y = np.argmax(y, axis=1)
self.categorical_features = [False] * num_features
self.problem_type = ProblemType.FeatureClassification
self.X, self.Y = X, Y
self._split_data(test_split, seed)
def test_shuffle_shuffles_splits(self):
X, y = make_multilabel_classification(100,
20,
n_labels=5,
random_state=0,
allow_unlabeled=False)
# With Shuffle
iskf = IterativeStratifiedKFold(n_splits=3,
shuffle=True,
random_state=42)
cv1 = list(iskf.split(X, y))
# Without shuffle
iskf = IterativeStratifiedKFold(n_splits=3,
shuffle=False,
random_state=42)
cv2 = list(iskf.split(X, y))
for train_shuff, valid_shuff in cv1:
for train_no_shuff, valid_no_shuff in cv2:
self.assertNotEqual(list(train_shuff), list(train_no_shuff))
self.assertNotEqual(list(valid_shuff), list(valid_no_shuff))
def test_pca_fit(datatype, input_type, name, use_handle):
if name == 'blobs':
pytest.skip('fails when using blobs dataset')
X, y = make_blobs(n_samples=500000, n_features=1000, random_state=0)
elif name == 'digits':
X, _ = datasets.load_digits(return_X_y=True)
else:
X, Y = make_multilabel_classification(n_samples=500,
n_classes=2,
n_labels=1,
allow_unlabeled=False,
random_state=1)
skpca = skPCA(n_components=2)
skpca.fit(X)
handle, stream = get_handle(use_handle)
cupca = cuPCA(n_components=2, handle=handle)
cupca.fit(X)
cupca.handle.sync()
for attr in [
'singular_values_', 'components_', 'explained_variance_',
'explained_variance_ratio_'
]:
with_sign = False if attr in ['components_'] else True
print(attr)
print(getattr(cupca, attr))
print(getattr(skpca, attr))
cuml_res = (getattr(cupca, attr))
if type(cuml_res) == np.ndarray:
cuml_res = cuml_res.as_matrix()
skl_res = getattr(skpca, attr)
assert array_equal(cuml_res, skl_res, 1e-3, with_sign=with_sign)
def __configure(self, n_samples, n_features, n_targets, n_labels):
""" __configure
Uses the make_multilabel_classification function from scikit-learn
to generate a multilabel classification problem. This problem will
be kept in memory and provided as demanded.
Parameters
----------
n_samples: int
Total amount of samples to generate.
n_features: int
Number of features to generate.
n_targets: int
Number of targeting tasks to generate.
n_labels: int
Number of labels to generate.
"""
self.X, self.y = make_multilabel_classification(
n_samples=n_samples,
n_features=n_features,
n_classes=n_targets,
n_labels=n_labels,
random_state=self.random_state)
self.num_samples = n_samples
self.num_features = n_features
self.num_target_tasks = n_targets
self.num_labels = n_labels
self.num_numerical_attributes = n_features
self.class_header = ["label_" + str(i) for i in range(self.num_labels)]
self.attributes_header = [
"att_num_" + str(i) for i in range(self.num_numerical_attributes)
]
def setup_mlc_dataset(self):
X, Y = datasets.make_multilabel_classification(n_features=5,
random_state=1126)
return Dataset(X, Y)
def check_classifier_on_multilabel_or_multioutput_targets(name, Estimator):
estimator = Estimator()
X, y = make_multilabel_classification(n_samples=30)
msg = "Multilabel and multioutput targets are not supported."
with pytest.raises(ValueError, match=msg):
estimator.fit(X, y)
def test_sparse_input(name, sparse_matrix):
X, y = datasets.make_multilabel_classification(random_state=0,
n_samples=50)
check_sparse_input(name, X, sparse_matrix(X), y)
def test_random_hasher_sparse_data():
X, y = datasets.make_multilabel_classification(random_state=0)
hasher = RandomTreesEmbedding(n_estimators=30, random_state=1)
X_transformed = hasher.fit_transform(X)
X_transformed_sparse = hasher.fit_transform(csc_matrix(X))
assert_array_equal(X_transformed_sparse.toarray(), X_transformed.toarray())
# # https://scikit-learn.org/stable/modules/generated/sklearn.decomposition.LatentDirichletAllocation.html # # # X contiene las noticias, pero quiero ver en qué formato # # In[1]: # Parallelizing using Pool.apply() import multiprocessing as mp from sklearn.decomposition import LatentDirichletAllocation from sklearn.datasets import make_multilabel_classification X, _ = make_multilabel_classification(random_state=0) print(X) print(X.shape) lda = LatentDirichletAllocation(n_components=5, random_state=0) lda.fit(X) lda.transform(X[-2:]) # Vale, parece que X es un array de numpy, que contiene 100 filas con 20 columnas cada uno: un vector fila por cada noticia, que contiene el conteo de veces que aparece cada palabra (de un total de 20 palabras). Esto es lo que se llama "Bag Of Words" # # En nuestro caso tenemos que coger todas las noticias, eliminar las palabras te tipo "stopwords" (palabras como "de", "la", "un", ...) y hacernos una lista con todas las palabras únicas que aparecen en el conjunto de todas las noticias (que serán más de 20 seguro, pero bueno). # ## Leer y limpiar csv # In[1]:
def dump_multilabel_classification(
model,
suffix="",
folder=None,
allow_failure=None,
verbose=False,
label_string=False,
first_class=0,
comparable_outputs=None,
target_opset=None):
"""
Trains and dumps a model for a binary classification problem.
The function trains a model and calls
:func:`dump_data_and_model`.
Every created filename will follow the pattern:
``<folder>/<prefix><task><classifier-name><suffix>.<data|expected|model|onnx>.<pkl|onnx>``.
"""
X = [[0, 1], [1, 1], [2, 0], [0.5, 0.5], [1.1, 1.1], [2.1, 0.1]]
X = numpy.array(X, dtype=numpy.float32)
if label_string:
y = [["l0"], ["l1"], ["l2"], ["l0", "l1"], ["l1"], ["l2"]]
else:
y = [[0 + first_class], [1 + first_class], [2 + first_class],
[0 + first_class, 1 + first_class],
[1 + first_class], [2 + first_class]]
y = MultiLabelBinarizer().fit_transform(y)
model.fit(X, y)
if verbose:
print("[make_multilabel_classification] model '{}'".format(
model.__class__.__name__))
model_onnx, prefix = convert_model(
model, "multi-class classifier",
[("input", FloatTensorType([None, 2]))],
target_opset=target_opset)
if verbose:
print("[make_multilabel_classification] model was converted")
dump_data_and_model(
X.astype(numpy.float32),
model,
model_onnx,
folder=folder,
allow_failure=allow_failure,
basename=prefix + "Mcl" + model.__class__.__name__ + suffix,
verbose=verbose,
comparable_outputs=comparable_outputs,
)
X, y = make_multilabel_classification(40, n_features=4, random_state=42,
n_classes=3)
X = X[:, :2]
model.fit(X, y)
if verbose:
print("[make_multilabel_classification] model '{}'".format(
model.__class__.__name__))
model_onnx, prefix = convert_model(model, "multi-class classifier",
[("input", FloatTensorType([None, 2]))])
if verbose:
print("[make_multilabel_classification] model was converted")
dump_data_and_model(
X[:10].astype(numpy.float32),
model,
model_onnx,
folder=folder,
allow_failure=allow_failure,
basename=prefix + "RndMla" + model.__class__.__name__ + suffix,
verbose=verbose,
comparable_outputs=comparable_outputs,
)
filename = input("Введите путь файла: ")
# Define the color maps for plots
color_map = plt.cm.get_cmap('RdYlBu')
color_map_discrete = matplotlib.colors.LinearSegmentedColormap.from_list("", ["red","cyan","magenta","blue"])
fig, ax = plt.subplots(nrows=1, ncols=4, figsize=(18, 7))
plt_ind_list = np.arange(3) + 131
dataset_x = []
dataset_y = []
dataset_sparse = []
labels = [1,2,4]
for label, plt_ind in zip(labels, plt_ind_list):
x, y = dt.make_multilabel_classification(n_samples=1000,
n_features=4,
n_labels=label,
n_classes=5,
random_state=rand_state)
target = np.sum(y * [1,1,1,1,1], axis=1)
dataset_x.append(x)
dataset_y.append(y)
plt.subplot(plt_ind)
my_scatter_plot = plt.scatter(x[:, 0],
x[:, 1],
c=target,
vmin=min(target),
vmax=max(target),
cmap=color_map)
plt.title('n_labels: ' + str(label))
n_ds_x = np.concatenate(dataset_x)
n_ds_y = np.concatenate(dataset_y)
