Fixing issues with imbalanced datasets #197
Merged
Conversation
This file contains 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
…ed because X was a numpy array at the time of checking
ravinkohli
suggested changes
May 5, 2021
There was a problem hiding this comment.
Hey Arlind, thanks for the PR. Its essential in the corner case when we have all nans in train but not in test and vice versa. But, could you add tests for the validator to ensure we can handle the case properly. It would be great if we could test the two cases- 1. where we have all NaNs in train and not in test and 2. where we have all NaNs in test but not in train
Yep, I will do just that today. But overall, I wanted to get a quick idea on how the problem is handled now, in case I missed something on the fix. |
… to check the implementation
…what happens at the validator, so the types do not change
ravinkohli
reviewed
May 7, 2021
| # for the test set, if we have columns with only null values | ||
| # they will probably have a numeric type. If these columns were not | ||
| # with only null values in the train set, they should be converted | ||
| # to the category type. |
There was a problem hiding this comment.
Suggested change
| # to the category type. | |
| # to the type it had while fitting. |
I don't think we should wait for the tests to merge in case you choose to incorporate this suggestion
There was a problem hiding this comment.
I also changed the comment accordingly.
ravinkohli
reviewed
May 7, 2021
| @@ -414,7 +422,7 @@ def infer_objects(self, X: pd.DataFrame) -> pd.DataFrame: | |||
| # In the case train data was interpreted as int | |||
| # and test data was interpreted as float, because of 0.0 | |||
| # for example, honor training data | |||
| X[key] = X[key].applymap(np.int64) | |||
| X[key] = X[key].astype(np.int64) | |||
There was a problem hiding this comment.
should we remove the if-else statement? as we are treating them equally
ravinkohli
reviewed
May 7, 2021
| if transformed_X_test[column].isna().all(): | ||
| null_columns.append(column) | ||
|
|
||
| assert null_columns == [1] |
There was a problem hiding this comment.
why is null_columns not empty here?
There was a problem hiding this comment.
Because for the test set, column 'A' will have only null values, however, it will have a numeric type. While column 'C' will have the 1 as a value through the entire column as it will detect correctly that it has 2 categories from the train set.
There was a problem hiding this comment.
The imputer is not used here for numerical values.
There was a problem hiding this comment.
I see, sorry I got confused as I expected none to be Null, I don't know why
ravinkohli
approved these changes
May 7, 2021
ravinkohli
pushed a commit
that referenced
this pull request
Dec 8, 2021
* adding missing method from base_feature_validator * First try at a fix, removing redundant code * Fix bug * Updating unit test typo, fixing bug where the data type was not checked because X was a numpy array at the time of checking * Fixing flake 8 failing * Bug fix, implementation update for imbalanced datasets and unit tests to check the implementation * flake8 fix * Bug fix * Making the conversion to dataframe in the unit tests consistent with what happens at the validator, so the types do not change * flake8 fix * Addressing Ravin's comments
ravinkohli
pushed a commit
that referenced
this pull request
Dec 8, 2021
* adding missing method from base_feature_validator * First try at a fix, removing redundant code * Fix bug * Updating unit test typo, fixing bug where the data type was not checked because X was a numpy array at the time of checking * Fixing flake 8 failing * Bug fix, implementation update for imbalanced datasets and unit tests to check the implementation * flake8 fix * Bug fix * Making the conversion to dataframe in the unit tests consistent with what happens at the validator, so the types do not change * flake8 fix * Addressing Ravin's comments
ravinkohli
pushed a commit
that referenced
this pull request
Dec 21, 2021
* adding missing method from base_feature_validator * First try at a fix, removing redundant code * Fix bug * Updating unit test typo, fixing bug where the data type was not checked because X was a numpy array at the time of checking * Fixing flake 8 failing * Bug fix, implementation update for imbalanced datasets and unit tests to check the implementation * flake8 fix * Bug fix * Making the conversion to dataframe in the unit tests consistent with what happens at the validator, so the types do not change * flake8 fix * Addressing Ravin's comments
ravinkohli
pushed a commit
that referenced
this pull request
Jan 24, 2022
* adding missing method from base_feature_validator * First try at a fix, removing redundant code * Fix bug * Updating unit test typo, fixing bug where the data type was not checked because X was a numpy array at the time of checking * Fixing flake 8 failing * Bug fix, implementation update for imbalanced datasets and unit tests to check the implementation * flake8 fix * Bug fix * Making the conversion to dataframe in the unit tests consistent with what happens at the validator, so the types do not change * flake8 fix * Addressing Ravin's comments
ravinkohli
pushed a commit
that referenced
this pull request
Jan 28, 2022
* adding missing method from base_feature_validator * First try at a fix, removing redundant code * Fix bug * Updating unit test typo, fixing bug where the data type was not checked because X was a numpy array at the time of checking * Fixing flake 8 failing * Bug fix, implementation update for imbalanced datasets and unit tests to check the implementation * flake8 fix * Bug fix * Making the conversion to dataframe in the unit tests consistent with what happens at the validator, so the types do not change * flake8 fix * Addressing Ravin's comments
ravinkohli
pushed a commit
that referenced
this pull request
Feb 28, 2022
* adding missing method from base_feature_validator * First try at a fix, removing redundant code * Fix bug * Updating unit test typo, fixing bug where the data type was not checked because X was a numpy array at the time of checking * Fixing flake 8 failing * Bug fix, implementation update for imbalanced datasets and unit tests to check the implementation * flake8 fix * Bug fix * Making the conversion to dataframe in the unit tests consistent with what happens at the validator, so the types do not change * flake8 fix * Addressing Ravin's comments
ravinkohli
pushed a commit
that referenced
this pull request
Feb 28, 2022
* adding missing method from base_feature_validator * First try at a fix, removing redundant code * Fix bug * Updating unit test typo, fixing bug where the data type was not checked because X was a numpy array at the time of checking * Fixing flake 8 failing * Bug fix, implementation update for imbalanced datasets and unit tests to check the implementation * flake8 fix * Bug fix * Making the conversion to dataframe in the unit tests consistent with what happens at the validator, so the types do not change * flake8 fix * Addressing Ravin's comments
ravinkohli
pushed a commit
that referenced
this pull request
Mar 9, 2022
* adding missing method from base_feature_validator * First try at a fix, removing redundant code * Fix bug * Updating unit test typo, fixing bug where the data type was not checked because X was a numpy array at the time of checking * Fixing flake 8 failing * Bug fix, implementation update for imbalanced datasets and unit tests to check the implementation * flake8 fix * Bug fix * Making the conversion to dataframe in the unit tests consistent with what happens at the validator, so the types do not change * flake8 fix * Addressing Ravin's comments
ravinkohli
pushed a commit
to ravinkohli/Auto-PyTorch
that referenced
this pull request
Apr 12, 2022
* adding missing method from base_feature_validator * First try at a fix, removing redundant code * Fix bug * Updating unit test typo, fixing bug where the data type was not checked because X was a numpy array at the time of checking * Fixing flake 8 failing * Bug fix, implementation update for imbalanced datasets and unit tests to check the implementation * flake8 fix * Bug fix * Making the conversion to dataframe in the unit tests consistent with what happens at the validator, so the types do not change * flake8 fix * Addressing Ravin's comments
ravinkohli
pushed a commit
that referenced
this pull request
Jul 26, 2022
* adding missing method from base_feature_validator * First try at a fix, removing redundant code * Fix bug * Updating unit test typo, fixing bug where the data type was not checked because X was a numpy array at the time of checking * Fixing flake 8 failing * Bug fix, implementation update for imbalanced datasets and unit tests to check the implementation * flake8 fix * Bug fix * Making the conversion to dataframe in the unit tests consistent with what happens at the validator, so the types do not change * flake8 fix * Addressing Ravin's comments
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.
This PR fixes issue #195