add inverse transform as final step in forecast method; also rephrase…

etna/pipeline/pipeline.py

@@ -106,6 +106,7 @@ def fit(self, ts: TSDataset) -> "Pipeline":
         self.ts = ts
         self.ts.fit_transform(self.transforms)
         self.model.fit(self.ts)
+        self.ts.inverse_transform()
         return self
 
     def _forecast_prediction_interval(self, future: TSDataset) -> TSDataset:
@@ -169,7 +170,7 @@ def _validate_backtest_n_folds(n_folds: int):
 
     @staticmethod
     def _validate_backtest_dataset(ts: TSDataset, n_folds: int, horizon: int):
-        """Check that all the given timestamps have enough timestamp points to validate forecaster with given number of splits."""
+        """Check all segments have enough timestamps to validate forecaster with given number of splits."""
         min_required_length = horizon * n_folds
         segments = set(ts.df.columns.get_level_values("segment"))
         for segment in segments:

tests/test_pipeline/test_pipeline.py

@@ -70,6 +70,7 @@ def test_fit(example_tsds):
     pipeline = Pipeline(model=model, transforms=transforms, horizon=5)
     pipeline.fit(example_tsds)
     original_ts.fit_transform(transforms)
+    original_ts.inverse_transform()
     assert np.all(original_ts.df.values == pipeline.ts.df.values)