[MRG] SimpleImputer: Handle string features where all values are missing #18860
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Ah shoot, let me fix these linting problems. |
thomasjpfan
reviewed
Nov 19, 2020
Comment on lines
+235
to
+236
| (isinstance(self.fill_value, str) | ||
| or any(isinstance(elem, str) for row in X for elem in row)): |
There was a problem hiding this comment.
This is using fill_value to determine the dtype of X and would cast X to an object dtype for input such as:
[[np.nan], [1], [np.nan]]I am unsure if we want to do this. The current code is considering [[np.nan], [np.nan]] a numerical feature and raising an error if the fill_value is not numerical, which looks to be correct behavior.
If one using None as the missing value, then everything still works:
from sklearn.impute import SimpleImputer
imputer = SimpleImputer(strategy='constant', fill_value='UNKNOWN', missing_values=None)
X_2 = [[None], [None]]
imputer.fit_transform(X_2)
# array([['UNKNOWN'],
# ['UNKNOWN']], dtype=object)Related to: #17625
thomasjpfan
reviewed
Nov 19, 2020
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Reference Issues/PRs
#17526 - This is relevant, as I don't think it was possible to run into this issue before.
What does this implement/fix? Explain your changes.
When dealing with a sparse string feature, it's not unlikely that a particular CV split has all missing values for said feature. In cases like this,
SimpleImputerappears to break. I believe this is because_validate_inputcurrently looks for astrvalue to determinedtype.To me, it seems that if a user specifies a string for
fill_value, that's another good indicator thatdtypeshould beobject.Any other comments?
Thanks for your time and consideration!