New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[BUG] FourierFeatures
with pd.DatetimeIndex
#3605
Comments
Thanks for spotting this @eenticott-shell. On a side note @fkiraly, isn't the "enforce_index_type" tag, supposed to enforce the index type? FYI - @topher-lo and @danbartl for our discussion on enforcing index types earlier. |
@ltsaprounis, the It does not coerce (which it probably should, from a design perspective) but instead acts as a "pass list": the boilerplate is supposed to raise an exception for index types not on the pass list - at least this is my understanding and what the contract description in the extension template says. If it lets other things pass (that are not subtypes), the actual logic does not agree with the contract, and it would be a bug. |
All is quantum physics |
OK - we have ourselves a bug then! |
🐛 😱 |
Describe the bug
When using data with a
pd.DatetimeIndex
in theFourierFeatures
transformer, it gets converted into nanoseconds. This means you have to specify all of your sp in terms of nanoseconds which is not clear.To Reproduce
Expected behavior
Would expect it to work from the frequency of the index in the
DatetimeIndex
case.Additional context
Have added functionality for the expected behaviour in #3606.
The text was updated successfully, but these errors were encountered: