You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Teach OutliersTransform to ignore holidays, which are not outliers in some sense.
Proposal
Add parameters ignore_flag_column: Optional[str] = None -- if set, then it should be column with binary flag of holiday (which might be created from the HolidaysTransform)
Update the transform method: now it should ignore the outliers on dates with true values in ignore_flag_column.
Update the fit method: now it should check the ignore_flag_column for existence and correctness of the format (binary)
Test cases
Transform works correctly with set ignore_flag_column -- it should leave the holidays intersecting with outliers
Calling fit with absence/incorrect formatted holidays_flag_column raises the corresponding error
Additional context
The text was updated successfully, but these errors were encountered:
🚀 Feature Request
Teach
OutliersTransform
to ignore holidays, which are not outliers in some sense.Proposal
ignore_flag_column: Optional[str] = None
-- if set, then it should be column with binary flag of holiday (which might be created from theHolidaysTransform
)transform
method: now it should ignore the outliers on dates with true values inignore_flag_column
.fit
method: now it should check theignore_flag_column
for existence and correctness of the format (binary)Test cases
ignore_flag_column
-- it should leave the holidays intersecting with outliersfit
with absence/incorrect formattedholidays_flag_column
raises the corresponding errorAdditional context
The text was updated successfully, but these errors were encountered: