[BUG] StackingEnsemble doesn't work with segment named target

[Mr-Geekman]

🐛 Bug Report

If dataset has a segment named target there will be problem on using StackingEnsemble with backtest.

Expected behavior

Everything works fine.

It can be fixed by setting level for renaming of columns in StackingEnsemble._make_features.

How To Reproduce

Script:

from sklearn.ensemble import GradientBoostingRegressor

from etna.datasets import TSDataset, generate_ar_df
from etna.transforms import LagTransform, DateFlagsTransform
from etna.pipeline import Pipeline
from etna.models import ProphetModel, CatBoostMultiSegmentModel
from etna.ensembles import StackingEnsemble
from etna.metrics import MAE, SMAPE

HORIZON = 7


def get_ts() -> TSDataset:
    df = generate_ar_df(periods=100, start_time="2020-01-01", n_segments=5, freq="D")
    df.loc[df["segment"] == "segment_0", "segment"] = "target"
    df_wide = TSDataset.to_dataset(df)
    ts = TSDataset(df=df_wide, freq="D")
    return ts


def main():
    ts = get_ts()

    df_melt = ts
    model = ProphetModel()
    catboost_transforms = [LagTransform(in_column="target", lags=list(range(7, 20))), DateFlagsTransform()]
    pipeline = Pipeline(model=model, transforms=[], horizon=HORIZON)
    multicb_pipeline = Pipeline(transforms=catboost_transforms, model=CatBoostMultiSegmentModel(), horizon=HORIZON)

    stacking_ensemble = StackingEnsemble(pipelines=[pipeline, multicb_pipeline],
                                         final_model=GradientBoostingRegressor(), n_folds=5,
                                         n_jobs=4)
    _ = stacking_ensemble.backtest(df_melt, metrics=[MAE(), SMAPE()], n_folds=3, aggregate_metrics=True, n_jobs=2)


if __name__ == "__main__":
    main()

Environment

No response

Additional context

No response

Checklist

• Bug appears at the latest library version