def test_standard_scaler():
input_df_train = pd.DataFrame({
"feat1_num": [1.0, 0.5, 100.0],
})
expected_output_train = pd.DataFrame({
"feat1_num": [-0.70175673, -0.71244338, 1.4142001],
})
input_df_test = pd.DataFrame({
"feat1_num": [2.0, 4.0, 8.0],
})
expected_output_test = pd.DataFrame({
"feat1_num": [-0.68038342, -0.63763682, -0.55214362],
})
pred_fn, train_result, log = standard_scaler(input_df_train, ["feat1_num"])
test_result = pred_fn(input_df_test)
assert_almost_equal(expected_output_train.values,
train_result.values,
decimal=5)
assert_almost_equal(test_result.values,
expected_output_test.values,
decimal=5)
def test_standard_scaler():
input_df_train = pd.DataFrame({"feat1_num": [1.0, 0.5, 100.0]})
expected_output_train = pd.DataFrame(
{"feat1_num": [-0.70175673, -0.71244338, 1.4142001]}
)
input_df_test = pd.DataFrame({"feat1_num": [2.0, 4.0, 8.0]})
expected_output_test = pd.DataFrame(
{"feat1_num": [-0.68038342, -0.63763682, -0.55214362]}
)
pred_fn1, data1, log = standard_scaler(input_df_train, ["feat1_num"])
pred_fn2, data2, log = standard_scaler(
input_df_train, ["feat1_num"], suffix="_suffix"
)
pred_fn3, data3, log = standard_scaler(
input_df_train, ["feat1_num"], prefix="prefix_"
)
pred_fn4, data4, log = standard_scaler(
input_df_train,
["feat1_num"],
columns_mapping={"feat1_num": "feat1_num_raw"},
)
assert_almost_equal(expected_output_train.values, data1.values, decimal=5)
assert_almost_equal(
expected_output_test.values, pred_fn1(input_df_test).values, decimal=5
)
assert_almost_equal(
pd.concat(
[
expected_output_train,
input_df_train[["feat1_num"]].copy().add_suffix("_suffix"),
],
axis=1,
).values,
data2.values,
decimal=5,
)
assert_almost_equal(
pd.concat(
[
expected_output_test,
input_df_test[["feat1_num"]].copy().add_suffix("_suffix"),
],
axis=1,
).values,
pred_fn2(input_df_test).values,
decimal=5,
)
assert_almost_equal(
pd.concat(
[
expected_output_train,
input_df_train[["feat1_num"]].copy().add_prefix("prefix_"),
],
axis=1,
).values,
data3.values,
decimal=5,
)
assert_almost_equal(
pd.concat(
[
expected_output_test,
input_df_test[["feat1_num"]].copy().add_prefix("prefix_"),
],
axis=1,
).values,
pred_fn3(input_df_test).values,
decimal=5,
)
assert_almost_equal(
pd.concat(
[
expected_output_train,
input_df_train[["feat1_num"]].copy().add_suffix("_raw"),
],
axis=1,
).values,
data4.values,
decimal=5,
)
assert_almost_equal(
pd.concat(
[
expected_output_test,
input_df_test[["feat1_num"]].copy().add_suffix("_raw"),
],
axis=1,
).values,
pred_fn4(input_df_test).values,
decimal=5,
)