Fix outliers transforms on future with gap

CHANGELOG.md

@@ -28,7 +28,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Fix inference tests on new segments for `DeepARModel` and `TFTModel` ([#1109](https://github.com/tinkoff-ai/etna/pull/1109))
 - Fix alignment during forecasting in new NNs, add validation of context size during forecasting in new NNs, add validation of batch in `MLPNet` ([#1108](https://github.com/tinkoff-ai/etna/pull/1108))
 - Fix `MeanSegmentEncoderTransform` to work with subset of segments and raise error on new segments ([#1104](https://github.com/tinkoff-ai/etna/pull/1104))
--
+- Fix outliers transforms on future with gap ([#1147](https://github.com/tinkoff-ai/etna/pull/1147))
 - Fix `SegmentEncoderTransform` to work with subset of segments and raise error on new segments ([#1103](https://github.com/tinkoff-ai/etna/pull/1103))
 - Fix `SklearnTransform` in per-segment mode to work on subset of segments and raise error on new segments ([#1107](https://github.com/tinkoff-ai/etna/pull/1107))
 - Fix `OutliersTransform` and its children to raise error on new segments ([#1139](https://github.com/tinkoff-ai/etna/pull/1139))

etna/transforms/outliers/base.py

@@ -103,7 +103,9 @@ def transform(self, df: pd.DataFrame) -> pd.DataFrame:
         segments = df.columns.get_level_values("segment").unique().tolist()
         self._validate_segments(segments)
         for segment in segments:
-            result_df.loc[self.outliers_timestamps[segment], pd.IndexSlice[segment, self.in_column]] = np.NaN
+            # to locate only present indices
+            segment_outliers_timestamps = result_df.index.intersection(self.outliers_timestamps[segment])
+            result_df.loc[segment_outliers_timestamps, pd.IndexSlice[segment, self.in_column]] = np.NaN
         return result_df
 
     def inverse_transform(self, df: pd.DataFrame) -> pd.DataFrame:

pyproject.toml

@@ -188,7 +188,6 @@ line_length = 120
 minversion = "6.0"
 doctest_optionflags = "NORMALIZE_WHITESPACE IGNORE_EXCEPTION_DETAIL NUMBER"
 filterwarnings = [
-    "error",
     "ignore: Torchmetrics v0.9 introduced a new argument class property called `full_state_update` that",
     "ignore: TSDataset freq can't be inferred",
     "ignore: test_size, test_start and test_end cannot be",

tests/test_transforms/test_inference/test_inverse_transform.py

@@ -1391,6 +1391,10 @@ def _test_inverse_transform_future_with_target(
                 "ts_to_fill",
                 {},
             ),
+            # outliers
+            (DensityOutliersTransform(in_column="target"), "ts_with_outliers", {}),
+            (MedianOutliersTransform(in_column="target"), "ts_with_outliers", {}),
+            (PredictionIntervalOutliersTransform(in_column="target", model=ProphetModel), "ts_with_outliers", {}),
             # timestamp
             (
                 DateFlagsTransform(out_column="res"),
@@ -1448,22 +1452,6 @@ def test_inverse_transform_future_with_target_fail_resample(
         ts = request.getfixturevalue(dataset_name)
         self._test_inverse_transform_future_with_target(ts, transform, expected_changes=expected_changes)
 
-    @to_be_fixed(raises=Exception)
-    @pytest.mark.parametrize(
-        "transform, dataset_name, expected_changes",
-        [
-            # outliers
-            (DensityOutliersTransform(in_column="target"), "ts_with_outliers", {}),
-            (MedianOutliersTransform(in_column="target"), "ts_with_outliers", {}),
-            (PredictionIntervalOutliersTransform(in_column="target", model=ProphetModel), "ts_with_outliers", {}),
-        ],
-    )
-    def test_inverse_transform_future_with_target_failed_error(
-        self, transform, dataset_name, expected_changes, request
-    ):
-        ts = request.getfixturevalue(dataset_name)
-        self._test_inverse_transform_future_with_target(ts, transform, expected_changes=expected_changes)
-
 
 class TestInverseTransformFutureWithoutTarget:
     """Test inverse transform on future dataset with unknown target.

tests/test_transforms/test_inference/test_transform.py

@@ -1255,6 +1255,10 @@ def _test_transform_future_with_target(self, ts, transform, expected_changes, ga
                 "ts_to_fill",
                 {},
             ),
+            # outliers
+            (DensityOutliersTransform(in_column="target"), "ts_with_outliers", {}),
+            (MedianOutliersTransform(in_column="target"), "ts_with_outliers", {}),
+            (PredictionIntervalOutliersTransform(in_column="target", model=ProphetModel), "ts_with_outliers", {}),
             # timestamp
             (
                 DateFlagsTransform(out_column="res"),
@@ -1279,20 +1283,6 @@ def test_transform_future_with_target(self, transform, dataset_name, expected_ch
         ts = request.getfixturevalue(dataset_name)
         self._test_transform_future_with_target(ts, transform, expected_changes=expected_changes)
 
-    @to_be_fixed(raises=Exception)
-    @pytest.mark.parametrize(
-        "transform, dataset_name, expected_changes",
-        [
-            # outliers
-            (DensityOutliersTransform(in_column="target"), "ts_with_outliers", {}),
-            (MedianOutliersTransform(in_column="target"), "ts_with_outliers", {}),
-            (PredictionIntervalOutliersTransform(in_column="target", model=ProphetModel), "ts_with_outliers", {}),
-        ],
-    )
-    def test_transform_future_with_target_failed_error(self, transform, dataset_name, expected_changes, request):
-        ts = request.getfixturevalue(dataset_name)
-        self._test_transform_future_with_target(ts, transform, expected_changes=expected_changes)
-
 
 class TestTransformFutureWithoutTarget:
     """Test transform on future dataset with unknown target.