def test_correct_lag_when_using_freq(df_time):
# Expected
date_time = [
pd.Timestamp("2020-05-15 12:00:00"),
pd.Timestamp("2020-05-15 12:15:00"),
pd.Timestamp("2020-05-15 12:30:00"),
pd.Timestamp("2020-05-15 12:45:00"),
pd.Timestamp("2020-05-15 13:00:00"),
]
expected_results = {
"ambient_temp": [31.31, 31.51, 32.15, 32.39, 32.62],
"module_temp": [49.18, 49.84, 52.35, 50.63, 49.61],
"irradiation": [0.51, 0.79, 0.65, 0.76, 0.42],
"color": ["blue"] * 5,
"ambient_temp_lag_1h": [np.nan, np.nan, np.nan, np.nan, 31.31],
"module_temp_lag_1h": [np.nan, np.nan, np.nan, np.nan, 49.18],
"irradiation_lag_1h": [np.nan, np.nan, np.nan, np.nan, 0.51],
"ambient_temp_lag_15min": [np.nan, 31.31, 31.51, 32.15, 32.39],
"module_temp_lag_15min": [np.nan, 49.18, 49.84, 52.35, 50.63],
"irradiation_lag_15min": [np.nan, 0.51, 0.79, 0.65, 0.76],
}
expected_results_df = pd.DataFrame(
data=expected_results,
index=date_time,
)
# When freq is a string
transformer = LagFeatures(freq="1h")
df_tr = transformer.fit_transform(df_time)
assert df_tr.head(5).equals(
expected_results_df.drop(
[
"ambient_temp_lag_15min",
"module_temp_lag_15min",
"irradiation_lag_15min",
],
axis=1,
))
# When freq is a list
transformer = LagFeatures(freq=["1h", "15min"])
df_tr = transformer.fit_transform(df_time)
assert df_tr.head(5).equals(expected_results_df)
# When drop original is True
transformer = LagFeatures(freq=["1h"], drop_original=True)
df_tr = transformer.fit_transform(df_time)
assert df_tr.head(5).equals(expected_results_df[[
"color", "ambient_temp_lag_1h", "module_temp_lag_1h",
"irradiation_lag_1h"
]])
def test_correct_lag_when_using_periods(df_time):
# Expected
date_time = [
pd.Timestamp("2020-05-15 12:00:00"),
pd.Timestamp("2020-05-15 12:15:00"),
pd.Timestamp("2020-05-15 12:30:00"),
pd.Timestamp("2020-05-15 12:45:00"),
pd.Timestamp("2020-05-15 13:00:00"),
]
expected_results = {
"ambient_temp": [31.31, 31.51, 32.15, 32.39, 32.62],
"module_temp": [49.18, 49.84, 52.35, 50.63, 49.61],
"irradiation": [0.51, 0.79, 0.65, 0.76, 0.42],
"color": ["blue"] * 5,
"ambient_temp_lag_3": [np.nan, np.nan, np.nan, 31.31, 31.51],
"module_temp_lag_3": [np.nan, np.nan, np.nan, 49.18, 49.84],
"ambient_temp_lag_2": [np.nan, np.nan, 31.31, 31.51, 32.15],
"module_temp_lag_2": [np.nan, np.nan, 49.18, 49.84, 52.35],
}
expected_results_df = pd.DataFrame(data=expected_results, index=date_time)
# When period is an int.
transformer = LagFeatures(variables=["ambient_temp", "module_temp"],
periods=3)
df_tr = transformer.fit_transform(df_time)
assert df_tr.head(5).equals(
expected_results_df.drop(["ambient_temp_lag_2", "module_temp_lag_2"],
axis=1))
# When period is list.
transformer = LagFeatures(variables=["ambient_temp", "module_temp"],
periods=[3, 2])
df_tr = transformer.fit_transform(df_time)
assert df_tr.head(5).equals(expected_results_df)
# When drop original is True
transformer = LagFeatures(variables=["ambient_temp", "module_temp"],
periods=[3, 2],
drop_original=True)
df_tr = transformer.fit_transform(df_time)
assert df_tr.head(5).equals(
expected_results_df.drop(["ambient_temp", "module_temp"], axis=1))
def test_sort_index(df_time):
X = df_time.copy()
# Shuffle dataframe
Xs = X.sample(len(df_time)).copy()
transformer = LagFeatures(sort_index=True)
X_tr = transformer.fit_transform(Xs)
A = X[transformer.variables_].iloc[0:4].values
B = X_tr[transformer.get_feature_names_out(
transformer.variables_)].iloc[1:5].values
assert (A == B).all()
transformer = LagFeatures(sort_index=False)
X_tr = transformer.fit_transform(Xs)
A = Xs[transformer.variables_].iloc[0:4].values
B = X_tr[transformer.get_feature_names_out(
transformer.variables_)].iloc[1:5].values
assert (A == B).all()
import numpy as np
import pandas as pd
import pytest
from sklearn.base import clone
from sklearn.utils.estimator_checks import check_estimator
from tests.estimator_checks.estimator_checks import check_feature_engine_estimator
from feature_engine.timeseries.forecasting import (
ExpandingWindowFeatures,
LagFeatures,
WindowFeatures,
)
_estimators = [
LagFeatures(missing_values="ignore"),
WindowFeatures(missing_values="ignore"),
ExpandingWindowFeatures(missing_values="ignore"),
]
@pytest.mark.parametrize("estimator", _estimators)
def test_check_estimator_from_sklearn(estimator):
return check_estimator(estimator)
@pytest.mark.parametrize("estimator", _estimators)
def test_check_estimator_from_feature_engine(estimator):
return check_feature_engine_estimator(estimator)
@pytest.mark.parametrize("estimator", _estimators)
SmartCorrelatedSelection(),
DropHighPSIFeatures(bins=5),
SelectByShuffling(LogisticRegression(max_iter=2, random_state=1),
scoring="accuracy"),
SelectBySingleFeaturePerformance(LogisticRegression(max_iter=2,
random_state=1),
scoring="accuracy"),
RecursiveFeatureAddition(LogisticRegression(max_iter=2, random_state=1),
scoring="accuracy"),
RecursiveFeatureElimination(
LogisticRegression(max_iter=2, random_state=1),
scoring="accuracy",
threshold=-100,
),
SelectByTargetMeanPerformance(scoring="roc_auc", bins=3, regression=False),
])
def test_sklearn_compatible_selectors(estimator, check):
check(estimator)
# wrappers
@parametrize_with_checks([SklearnTransformerWrapper(SimpleImputer())])
def test_sklearn_compatible_wrapper(estimator, check):
check(estimator)
# test_forecasting
@parametrize_with_checks([LagFeatures(missing_values="ignore")])
def test_sklearn_compatible_forecasters(estimator, check):
check(estimator)
def test_permitted_param_periods(_periods):
transformer = LagFeatures(periods=_periods)
assert transformer.periods == _periods
def test_get_feature_names_out(df_time):
# input features
input_features = ["ambient_temp", "module_temp", "irradiation"]
original_features = ["ambient_temp", "module_temp", "irradiation", "color"]
# When freq is a string:
tr = LagFeatures(freq="1D")
tr.fit(df_time)
# Expected
out = ["ambient_temp_lag_1D", "module_temp_lag_1D", "irradiation_lag_1D"]
assert tr.get_feature_names_out(
input_features=None) == original_features + out
assert tr.get_feature_names_out(input_features=input_features) == out
assert tr.get_feature_names_out(
input_features=input_features[0:1]) == out[0:1]
assert tr.get_feature_names_out(input_features=[input_features[0]]) == [
out[0]
]
with pytest.raises(ValueError):
# assert error when user passes a string instead of list
tr.get_feature_names_out(input_features=input_features[0])
with pytest.raises(ValueError):
# assert error when uses passes features that were not lagged
tr.get_feature_names_out(input_features=["color"])
# When period is an int:
tr = LagFeatures(periods=2)
tr.fit(df_time)
# Expected
out = ["ambient_temp_lag_2", "module_temp_lag_2", "irradiation_lag_2"]
assert tr.get_feature_names_out(
input_features=None) == original_features + out
assert tr.get_feature_names_out(input_features=input_features) == out
assert tr.get_feature_names_out(
input_features=input_features[0:1]) == out[0:1]
assert tr.get_feature_names_out(input_features=[input_features[0]]) == [
out[0]
]
# When freq is a list:
tr = LagFeatures(freq=["3D", "2D"])
tr.fit(df_time)
# Expected
out = [
"ambient_temp_lag_3D",
"module_temp_lag_3D",
"irradiation_lag_3D",
"ambient_temp_lag_2D",
"module_temp_lag_2D",
"irradiation_lag_2D",
]
assert tr.get_feature_names_out(
input_features=None) == original_features + out
assert tr.get_feature_names_out(input_features=input_features) == out
assert (tr.get_feature_names_out(
input_features=input_features[0:1]) == out[0:1] + out[3:4])
assert tr.get_feature_names_out(
input_features=[input_features[0]]) == [out[0]] + [out[3]]
# When periods is a list:
tr = LagFeatures(periods=[2, 3])
tr.fit(df_time)
# Expected
out = [
"ambient_temp_lag_2",
"module_temp_lag_2",
"irradiation_lag_2",
"ambient_temp_lag_3",
"module_temp_lag_3",
"irradiation_lag_3",
]
assert tr.get_feature_names_out(
input_features=None) == original_features + out
assert tr.get_feature_names_out(input_features=input_features) == out
assert (tr.get_feature_names_out(
input_features=input_features[0:1]) == out[0:1] + out[3:4])
assert tr.get_feature_names_out(
input_features=[input_features[0]]) == [out[0]] + [out[3]]
# When drop original is true.
tr = LagFeatures(freq="1D", drop_original=True)
tr.fit(df_time)
# Expected
out = ["ambient_temp_lag_1D", "module_temp_lag_1D", "irradiation_lag_1D"]
assert tr.get_feature_names_out(input_features=None) == ["color"] + out
assert tr.get_feature_names_out(input_features=input_features) == out
def test_error_when_non_permitted_param_sort_index(_sort_index):
with pytest.raises(ValueError):
LagFeatures(sort_index=_sort_index)
def test_permitted_param_sort_index(_sort_index):
transformer = LagFeatures(sort_index=_sort_index)
assert transformer.sort_index == _sort_index
def test_error_when_non_permitted_param_freq():
with pytest.raises(ValueError):
LagFeatures(freq=["2h", "2h", "3h"])
def test_error_when_non_permitted_param_periods(_periods):
with pytest.raises(ValueError):
LagFeatures(periods=_periods)