def test_fit_complex(complex_data_split, complex_fitted_lightgbm):
X_train, X_test, y_train, y_test = complex_data_split
plotter = TreeDependencePlotter(complex_fitted_lightgbm)
plotter.fit(X_test, y_test)
pd.testing.assert_frame_equal(plotter.X, X_test)
pd.testing.assert_series_equal(plotter.y, pd.Series(y_test, index=X_test.index))
assert plotter.fitted is True
# Check if plotting doesnt cause errors
with patch('matplotlib.pyplot.figure') as mock_plt:
for binning in ["simple", "agglomerative", "quantile"]:
_ = plotter.plot(feature='f2_missing', type_binning=binning)
def test_plot_input(X_y, clf):
plotter = TreeDependencePlotter(clf).fit(X_y[0], X_y[1])
with pytest.raises(ValueError):
plotter.plot(feature="not a feature")
with pytest.raises(ValueError):
plotter.plot(feature=0, type_binning=5)
with pytest.raises(ValueError):
plotter.plot(feature=0, min_q=1, max_q=0)
def test_get_X_y_shap_with_q_cut_normal(X_y, clf):
X, y = X_y
plotter = TreeDependencePlotter(clf).fit(X, y)
plotter.min_q, plotter.max_q = 0, 1
X_cut, y_cut, shap_val = plotter._get_X_y_shap_with_q_cut(0)
assert np.isclose(X[0], X_cut).all()
assert y.equals(y_cut)
plotter.min_q = 0.2
plotter.max_q = 0.8
X_cut, y_cut, shap_val = plotter._get_X_y_shap_with_q_cut(0)
assert np.isclose(
X_cut,
[
-1.48382902,
-0.44947744,
-1.38101231,
-0.18261804,
0.27514902,
-0.27264455,
-1.27251335,
-2.10917352,
-1.25945582,
],
).all()
assert np.equal(y_cut.values, [1, 0, 0, 1, 1, 0, 0, 0, 0]).all()
def test_not_fitted(clf):
plotter = TreeDependencePlotter(clf)
assert plotter.fitted is False
def test__repr__(clf):
plotter = TreeDependencePlotter(clf)
assert str(plotter) == "Shap dependence plotter for RandomForestClassifier"
def test_plot_class_names(X_y, clf):
plotter = TreeDependencePlotter(clf).fit(X_y[0], X_y[1], class_names=["a", "b"])
fig = plotter.plot(feature=0)
assert plotter.class_names == ["a", "b"]
def test_plot_normal(X_y, clf):
plotter = TreeDependencePlotter(clf).fit(X_y[0], X_y[1])
for binning in ["simple", "agglomerative", "quantile"]:
fig = plotter.plot(feature=0, type_binning=binning)
def test_get_X_y_shap_with_q_cut_input(X_y, clf):
plotter = TreeDependencePlotter(clf).fit(X_y[0], X_y[1])
with pytest.raises(ValueError):
plotter._get_X_y_shap_with_q_cut("not a feature")
def test_get_X_y_shap_with_q_cut_unfitted(clf):
plotter = TreeDependencePlotter(clf)
with pytest.raises(NotFittedError):
plotter._get_X_y_shap_with_q_cut(0)