[SPARK-22734][ML][PySpark] Added Python API for VectorSizeHint. #20112
Conversation
|
Test build #85496 has finished for PR 20112 at commit
|
There was a problem hiding this comment.
Thanks for the PR!
You'll need to add this class to the __all__ list at the top of the file.
Since VectorSizeHint modifies the inputCol, rather than creating a new outputCol, we should explicitly document this. Would you mind adding a note to the class doc string in both Python and Scala? (It's OK to do both in this PR.)
python/pyspark/ml/feature.py
Outdated
| self._set(size=value) | ||
|
|
||
| @keyword_only | ||
| def __init__(self, inputCol=None, size=None, handleInvalid="error"): |
There was a problem hiding this comment.
Let's stick with the order which all other python classes follow: dummy Params, init, Param setters & getters
| @@ -3466,6 +3466,72 @@ def selectedFeatures(self): | |||
| return self._call_java("selectedFeatures") | |||
|
|
|||
|
|
|||
| @inherit_doc | |||
| class VectorSizeHint(JavaTransformer, HasInputCol, HasHandleInvalid, JavaMLReadable, | |||
There was a problem hiding this comment.
You'll need to override handleInvalid, like in the Scala API, since it takes different values & has a different docstring.
MrBago
force-pushed
the
vectorSizeHint-PythonAPI
branch
from
9a1ee2b
to
1ec7a41
Compare
|
Test build #85527 has finished for PR 20112 at commit
|
|
Test build #85528 has finished for PR 20112 at commit
|
|
Looks good except for the style issue |
|
Test build #85537 has finished for PR 20112 at commit
|
|
LGTM |
(Please fill in changes proposed in this fix)
Python API for VectorSizeHint Transformer.
(Please explain how this patch was tested. E.g. unit tests, integration tests, manual tests)
doc-tests.