def test_extract_datetime_features_with_default_options(
df_datetime, df_datetime_transformed):
transformer = DatetimeFeatures()
X = transformer.fit_transform(df_datetime)
pd.testing.assert_frame_equal(
X,
df_datetime_transformed[
vars_non_dt +
[var + feat for var in vars_dt for feat in feat_names_default]],
)
def test_extract_features_from_different_timezones(df_datetime,
df_datetime_transformed):
time_zones = [4, -1, 9, -7]
tz_df = pd.DataFrame(
{"time_obj": df_datetime["time_obj"].add(["+4", "-1", "+9", "-7"])})
transformer = DatetimeFeatures(variables="time_obj",
features_to_extract=["hour"],
utc=True)
X = transformer.fit_transform(tz_df)
pd.testing.assert_frame_equal(
X,
df_datetime_transformed[["time_obj_hour"
]].apply(lambda x: x.subtract(time_zones)),
)
exp_err_msg = ("ValueError: variable(s) time_obj "
"could not be converted to datetime. Try setting utc=True")
with pytest.raises(ValueError) as errinfo:
assert DatetimeFeatures(variables="time_obj",
features_to_extract=["hour"],
utc=False).fit_transform(tz_df)
assert str(errinfo.value) == exp_err_msg
def test_get_feature_names_out(df_datetime, df_datetime_transformed):
# default features from all variables
transformer = DatetimeFeatures()
X = transformer.fit_transform(df_datetime)
assert list(X.columns) == transformer.get_feature_names_out()
assert transformer.get_feature_names_out(input_features=vars_dt) == [
var + feat for var in vars_dt for feat in feat_names_default
]
assert transformer.get_feature_names_out(input_features=["date_obj1"]) == [
"date_obj1" + feat for feat in feat_names_default
]
# default features from 1 variable
transformer = DatetimeFeatures(variables="date_obj1")
X = transformer.fit_transform(df_datetime)
assert list(X.columns) == transformer.get_feature_names_out()
assert transformer.get_feature_names_out(input_features=["date_obj1"]) == [
"date_obj1" + feat for feat in feat_names_default
]
# all features
transformer = DatetimeFeatures(features_to_extract="all")
X = transformer.fit_transform(df_datetime)
assert list(X.columns) == transformer.get_feature_names_out()
feat_names_all = [FEATURES_SUFFIXES[feat] for feat in FEATURES_SUPPORTED]
assert transformer.get_feature_names_out(input_features=vars_dt) == [
var + feat for var in vars_dt for feat in feat_names_all
]
assert transformer.get_feature_names_out(input_features=["date_obj1"]) == [
"date_obj1" + feat for feat in feat_names_all
]
# specified features
transformer = DatetimeFeatures(features_to_extract=["semester", "week"])
X = transformer.fit_transform(df_datetime)
assert list(X.columns) == transformer.get_feature_names_out()
assert transformer.get_feature_names_out(input_features=vars_dt) == [
var + "_" + feat for var in vars_dt for feat in ["semester", "week"]
]
assert transformer.get_feature_names_out(input_features=["date_obj1"]) == [
"date_obj1_" + feat for feat in ["semester", "week"]
]
# features were extracted from index
transformer = DatetimeFeatures(variables="index",
features_to_extract=["semester", "week"])
X = transformer.fit_transform(dates_idx_dt)
# user passes something else than index as input_features
with pytest.raises(ValueError):
transformer.get_feature_names_out(input_features="not_index")
with pytest.raises(ValueError):
transformer.get_feature_names_out(
input_features=["still", "not", "index"])
# input_features is None
assert list(X.columns) == transformer.get_feature_names_out()
# input_features is index
assert ["semester", "week"
] == transformer.get_feature_names_out(input_features="index")
# when drop original is False
transformer = DatetimeFeatures(drop_original=False)
X = transformer.fit_transform(df_datetime)
assert list(X.columns) == transformer.get_feature_names_out()
assert transformer.get_feature_names_out(input_features=vars_dt) == [
var + feat for var in vars_dt for feat in feat_names_default
]
assert transformer.get_feature_names_out(input_features=["date_obj1"]) == [
"date_obj1" + feat for feat in feat_names_default
]
with pytest.raises(ValueError):
# assert error when user passes a string instead of list
transformer.get_feature_names_out(input_features="date_obj1")
with pytest.raises(ValueError):
# assert error when uses passes features that were not lagged
transformer.get_feature_names_out(input_features=["color"])