search

paris-saclay-cds / ramp-workflow

Toolkit for building predictive workflows on top of pydata (pandas, scikit-learn, pytorch, keras, etc.).
https://paris-saclay-cds.github.io/ramp-docs/
BSD 3-Clause "New" or "Revised" License
68 stars 43 forks source link

time series #315

Open kegl opened 1 year ago

kegl commented 1 year ago

We have decided to cherry pick our generic time series features from our generative regression branch and add it to RAMP in a separate PR. We'll also have new tests (get rid of El Nino and netcdf), and also a generic kit where we add popular time series predictors in individual submissions. Anybody using the current time series ingredients, eg https://github.com/paris-saclay-cds/ramp-workflow/blob/master/rampwf/cvs/time_series.py? Do you mind if we break El Nino (but provide a cleaner kit for time series)?

@albertcthomas @rth @agramfort @SofianChay

agramfort commented 1 year ago

no objection. El nino is not used anywhere at this point

Message ID: @.***>

albertcthomas commented 1 year ago

Should we pin the current rampwf version in the El Nino kit? although I am not even sure that it is still working with the current version.

kegl commented 1 year ago

No, we don't need to do that.