Make covariate column names explicit#34
Conversation
abdulfatir
left a comment
abdulfatir
left a comment
There was a problem hiding this comment.
Thanks @shchur! This will make the task definitions more transparent and easy to check which columns are actually used by a model.
| ) | ||
| future_known = future_data.remove_columns(self.target_columns_list + self.past_dynamic_columns) | ||
| test = future_data.select_columns([self.id_column, self.timestamp_column] + self.target_columns_list) | ||
| future_known = future_data.remove_columns(self.target_columns + self.past_dynamic_columns) |
There was a problem hiding this comment.
Should we specifically only select known future columns (+ id and timestamp) here after removing past and target_columns?
There was a problem hiding this comment.
It looks like all other columns are removed during data loading, so maybe this is not needed?
There was a problem hiding this comment.
I have updated load_full_dataset to not remove any columns, and to only perform column removal before the split. This makes it easy to check what columns are available in the dataset using public API.
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
2 participants
Issue #, if available:
Description of changes:
__version__stored in__init__.pyknown_dynamic_columns,past_dynamic_columns,static_columnsexplicit properties of theTaskinstead of automatically inferring them from data.target_column->targetto make the name more consistent withstr | list[str]typing.Task.target_columns_list->target_columns.To do
docs/By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.