def test_default_model():
df = generate_unstructured_test_data(1000)
df_checkpoint = df.copy()
features = ["A", "B", "C", "D", "E"]
lofo = LOFOImportance(df,
features,
'target',
cv=4,
scoring='neg_mean_absolute_error')
importance_df = lofo.get_importance()
assert len(features) == importance_df.shape[
0], "Missing importance value for some features!"
lofo = LOFOImportance(df,
features,
'binary_target',
cv=4,
scoring='roc_auc')
importance_df = lofo.get_importance()
assert df.equals(df_checkpoint), "LOFOImportance mutated the dataframe!"
assert importance_df["feature"].values[
0] == "E", "Most important feature is different than E!"
def test_default_model():
df = generate_unstructured_test_data(1000)
features = ["A", "B", "C", "D", "E"]
dataset = Dataset(df=df, target="target", features=features)
lofo = LOFOImportance(dataset, cv=4, scoring='neg_mean_absolute_error')
importance_df = lofo.get_importance()
assert "E" in lofo.fit_params[
"categorical_feature"], "Categorical feature is not detected!"
assert len(features) == importance_df.shape[
0], "Missing importance value for some features!"
df_checkpoint = df.copy()
dataset = Dataset(df=df, target="binary_target", features=features)
lofo = LOFOImportance(dataset, cv=4, scoring='roc_auc')
importance_df = lofo.get_importance()
assert "E" in lofo.fit_params[
"categorical_feature"], "Categorical feature is not detected!"
assert df.equals(df_checkpoint), "LOFOImportance mutated the dataframe!"
assert importance_df["feature"].values[
0] == "E", "Most important feature is different than E!"
def test_dataset():
df = generate_unstructured_test_data(1000, text=True)
features = ["A", "B", "C", "D", "D2", "E"]
# Exception: feature group row count is not equal to the features' row count
feature_groups = {
"interactions": df[["A", "B"]].values[:10] * df[["C", "D"]].values[:10]
}
with pytest.raises(Exception):
assert Dataset(df=df,
target="binary_target",
features=features,
feature_groups=feature_groups)
# Exception: Feature group name A is in use by other features
feature_groups = {"A": df[["A", "B"]].values * df[["C", "D"]].values}
with pytest.raises(Exception):
assert Dataset(df=df,
target="binary_target",
features=features,
feature_groups=feature_groups)
# Exception: Feature group type is not numpy.ndarray or scipy.csr.csr_matrix
feature_groups = {"F": df[["A", "B"]]}
with pytest.raises(Exception):
assert Dataset(df=df,
target="binary_target",
features=features,
feature_groups=feature_groups)
d = Dataset(df=df,
target="binary_target",
features=features,
feature_groups={"F": df[["A", "B"]].values},
auto_group_threshold=0.5)
assert "D" not in d.feature_names and "D2" not in d.feature_names
assert "D & D2" in d.feature_names and "F" in d.feature_groups.keys()