def test_interleaved_scores():
"""Test a dummy regressor"""
X = np.random.random((100, 10))
y = X[:, 0]**2
estimator = TemplateEstimator()
mse_scorer = make_scorer(mean_squared_error)
scores = interleaved_evaluation(estimator, X, y, mse_scorer)
assert_almost_equal(scores, np.zeros(X.shape[0]))
def test_template_estimator(data):
est = TemplateEstimator()
assert est.demo_param == 'demo_param'
est.fit(*data)
assert hasattr(est, 'is_fitted_')
X = data[0]
y_pred = est.predict(X)
assert_array_equal(y_pred, np.ones(X.shape[0], dtype=np.int64))
def test_prequential_window():
"""Test length of produced evaluation when windows are used"""
n_samples = 100
window_size = 10
assert n_samples % window_size == 0 # Meta-test, this needs to hold
X = np.random.random((n_samples, 10))
y = X[:, 0]**2
estimator = TemplateEstimator()
mse_scorer = make_scorer(mean_squared_error)
scores = prequential_evaluation(estimator,
X,
y,
mse_scorer,
window_size=window_size)
eq_(len(scores), n_samples / window_size)
def test_prequential_incomplete_window():
"""Test length of produced evaluation when windows are used, and the last
window has less items than the requested length"""
n_samples = 109
window_size = 10
assert n_samples % window_size != 0
X = np.random.random((n_samples, 10))
y = X[:, 0]**2
estimator = TemplateEstimator()
mse_scorer = make_scorer(mean_squared_error)
scores = prequential_evaluation(estimator,
X,
y,
mse_scorer,
window_size=window_size)
eq_(len(scores), np.ceil(n_samples / window_size))
def test_prequential_multimetric():
n_samples = 100
window_size = 10
assert n_samples % window_size == 0 # Meta-test, this needs to hold
X = np.random.random((n_samples, 10))
y = X[:, 0]**2
estimator = TemplateEstimator()
mse_scorer = make_scorer(mean_squared_error)
r2_scorer = make_scorer(r2_score)
score_dict = {"mse": mse_scorer, "r2": r2_scorer}
result_dict = prequential_evaluation(estimator,
X,
y,
score_dict,
window_size=window_size)
assert isinstance(result_dict, dict)
for score_name, score_list in result_dict.items():
eq_(len(score_list), n_samples / window_size)
""" =========================== Plotting Template Estimator =========================== An example plot of TemplateEstimator """ import numpy as np from skltemplate import TemplateEstimator from matplotlib import pyplot as plt X = np.arange(100).reshape(100, 1) y = np.zeros((100, )) estimator = TemplateEstimator() estimator.fit(X, y) plt.plot(estimator.predict(X)) plt.show()
import pytest
from sklearn.utils.estimator_checks import check_estimator
from skltemplate import TemplateEstimator
from skltemplate import TemplateClassifier
from skltemplate import TemplateTransformer
@pytest.mark.parametrize(
"estimator",
[TemplateEstimator(), TemplateTransformer(), TemplateClassifier()]
)
def test_all_estimators(estimator):
return check_estimator(estimator)
