🔮〰️ Update Prediction Filtering #1048
Conversation
…en/pykeen into update-prediction-filtering
|
|
||
| def filter(self, df: pd.DataFrame) -> pd.DataFrame: | ||
| """ | ||
| Filter out known triples. |
There was a problem hiding this comment.
should mention in docstring this is not an in-place operation, maybe should give the possibility for both ways?
There was a problem hiding this comment.
So far, we do not have a use-case for in-place operations, so I would defer the implementation until then 😉
| ), | ||
| test_elements=mapped_triples, | ||
| ) | ||
| .cpu() |
There was a problem hiding this comment.
maybe better to return the torch tensor then just cpu().numpy() in the place where it gets used?
There was a problem hiding this comment.
Currently, we need cpu().numpy() everytime we call this function, so I would leave it as it is for now.
|
Besides minor code comments I think this is fine, if it matches existing functionality then great. Will you later add some more high-level documentation when you're ready to expose to the public? |
Pull request was converted to draft
This (draft) PR introduces support for more complex prediction filtering, e.g., with multiple filter triples sets.
It does not (yet) expose this functionality to public.
related to #1040