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
Is your feature request related to a problem? Please describe.
For example, you want to create lag features of sales for a retail chain, but taking into account the stores, product categories, and date. It's widespread to create lag features taking into consideration the hierarchy or grouping levels.
Describe the solution you'd like
I would like the la feature creation but apply groups by different columns in the data frame.
Describe alternatives you've considered
An alternative solution would be to create a sklearn transformer class to apply the groupby and lag functions of pandas.
I have time series identified by an id column, and I would also need to create lag features for each group id.
Moreover, in addition to the id column, it would be useful to be able to pass the name of a column to use for sorting within each group (e.g., a datetime column) instead of just having the "sort_index" parameter.
Is your feature request related to a problem? Please describe.
For example, you want to create lag features of sales for a retail chain, but taking into account the stores, product categories, and date. It's widespread to create lag features taking into consideration the hierarchy or grouping levels.
Describe the solution you'd like
I would like the la feature creation but apply groups by different columns in the data frame.
Describe alternatives you've considered
An alternative solution would be to create a sklearn transformer class to apply the groupby and lag functions of pandas.
Here is an example of what I'm talking about using pandas: https://www.statology.org/pandas-lag-by-group/
The text was updated successfully, but these errors were encountered: