alex-hse-repository comments

Results 12 comments of


                                            alex-hse-repository

ChangePointSegmentationTransform

1. Create the base class for change points models `BaseChangePointModelAdapter`: - `get_change_points(self, df:pd DataFrame, in_column: str) -> List[pd.Timestamp]` -- abstract - `build_intervals(change_points: List[pd.Timestamp]) ->List[TTimestampInterval]` -- use the current implementation 2....

[DRAFT]Make base class of Transforms work with TSDataset

Special in_column handling in ResamplingTransform

[BUG] Dropping NaNs on fit in all of the models

Waiting for tests

Speed up columns slices: `etna.datasets.utils.select_columns`

Make sure that you do not forget to fix [this](https://github.com/tinkoff-ai/etna/blob/9e13e55ceddec45c2e5176d42c0b7e7651515b1c/etna/datasets/tsdataset.py#L886), [this](https://github.com/tinkoff-ai/etna/blob/fb09c5b9546a5f62b7fd090d70a2cdea6b4711b0/etna/datasets/tsdataset.py#L650) and [this](https://github.com/tinkoff-ai/etna/blob/fb09c5b9546a5f62b7fd090d70a2cdea6b4711b0/etna/datasets/tsdataset.py#L576) places in TSDataset

Speed up columns slices: `etna.datasets.utils.select_columns`

And [here](https://github.com/tinkoff-ai/etna/blob/6856c484be197d5b10994de4808502204cd41fda/etna/ensembles/direct_ensemble.py#L115)

Speed up columns slices: `etna.datasets.utils.select_columns`

And [here](https://github.com/tinkoff-ai/etna/blob/88b027cb8899055b44fef8a17ad59836068997e7/etna/datasets/tsdataset.py#L887), [here](https://github.com/tinkoff-ai/etna/blob/88b027cb8899055b44fef8a17ad59836068997e7/etna/transforms/base.py#L66)

get_anomalies_confidence_interval should expect Pipeline

Waiting for `in_sample` predictions in Pipeline

Should we allow TSDataset to hold non-continuous series?

Will work on it in the separate track

Feature selection

But we can leave this version as well --- View entire conversation on ReviewNB

Feature selection

May you check the results with fast_redundancy=False --- View entire conversation on ReviewNB