Add public APIs to Access Underlying cudf and pandas Objects from cudf.pandas Proxy Objects
#17629
Conversation
galipremsagar
added
improvement
github-actions
bot
added
Python
There was a problem hiding this comment.
Overall, this looks good to me. But I think we should add a bullet to the "Are there any limitations?" section of this faq.md . And it should describe the implications for users of cudf.pandas and third-party libraries that are "cudf aware." For example, they could get a cupy array (not a numpy array) when working with xgboost. What do you think?
Co-authored-by: Bradley Dice <[email protected]>
| and (index is None or index_extracted) | ||
| and (columns is None or columns_extracted) | ||
| ) and (dtype is None and copy is None): | ||
| self.__dict__.update(data.__dict__) |
There was a problem hiding this comment.
Could we use _mimic_inplace instead?
| @@ -451,3 +451,17 @@ def _datetime_timedelta_find_and_replace( | |||
| except TypeError: | |||
| result_col = original_column.copy(deep=True) | |||
| return result_col # type: ignore | |||
|
|
|||
|
|
|||
| def _extract_from_proxy(proxy, fast=True): | |||
There was a problem hiding this comment.
| def _extract_from_proxy(proxy, fast=True): | |
| def _extract_from_proxy(proxy: Any, fast: bool=True) -> tuple[Any, bool]: |
| def _Series_dtype(self): | ||
| # Fast-path to extract dtype from the current | ||
| # object without round-tripping through the slow<->fast | ||
| return self._fsproxy_wrapped.dtype |
There was a problem hiding this comment.
I'm a little nervous to start doing this because we need to be sure the dtype between the fast and slow object are equal correct? e.g. If _fsproxy_wrapped is a pandas.Series with a the extension pandas.Float64Dtype, would that break anything?
Description
Fixes: #17524
Fixes: rapidsai/cuml#6232
This PR introduces methods to access the real underlying
cudfandpandasobjects fromcudf.pandasproxy objects. These methods ensure compatibility with libraries that arecudforpandasaware.This PR also gives a performance boost to
cudf-pandasworkflows, speeds from the script posted in rapidsai/cuml#6232:branch-25.02:This PR:Changes:
get_gpu_object()andget_cpu_object()methods.Checklist