Fix test_describe_pipeline for pandas 1.2.4
#2129
Conversation
Codecov Report
@@ Coverage Diff @@
## main #2129 +/- ##
=======================================
Coverage 100.0% 100.0%
=======================================
Files 291 291
Lines 23809 23809
=======================================
Hits 23799 23799
Misses 10 10
Continue to review full report at Codecov.
|
angela97lin
requested review from
ParthivNaresh,
bchen1116,
chukarsten,
dsherry,
freddyaboulton and
jeremyliweishih
freddyaboulton
left a comment
freddyaboulton
left a comment
There was a problem hiding this comment.
Thanks for the fix @angela97lin !
I think this is the smallest change we can make to fix our tests but I think its weird we're displaying the dataset size to three decimal places since it should always be an int. In "old pandas" we would display to one decimal place which is better but I still think its weird.
I wonder if the better thing to do is either display the fold size outside of the dataframe or round to ints?
|
@freddyaboulton Yeah, I think the logic here is frustrating because we're converting to One option is to use the nullable Int64--that can be okay for this internal structure that isn't expected to be modified by the user. Also posting this: https://stackoverflow.com/questions/39690742/convert-float-to-int-and-leave-nulls, will try to see if any of these solutions float our boat :) EDIT: Found a way to display "ints" aka floats with less precision 😎 waiting for tests to pass, then will merge! |
| assert "0 1.000 66.000 34.000" in out | ||
| assert "1 1.000 67.000 33.000" in out | ||
| assert "2 1.000 67.000 33.000" in out | ||
| assert "0 1.000 66 34" in out |
…skEngine`` #1975. - Added optional ``engine`` argument to ``AutoMLSearch`` #1975 - Added a warning about how time series support is still in beta when a user passes in a time series problem to ``AutoMLSearch`` #2118 - Added ``NaturalLanguageNaNDataCheck`` data check #2122 - Added ValueError to ``partial_dependence`` to prevent users from computing partial dependence on columns with all NaNs #2120 - Added standard deviation of cv scores to rankings table #2154 - Fixed ``BalancedClassificationDataCVSplit``, ``BalancedClassificationDataTVSplit``, and ``BalancedClassificationSampler`` to use ``minority:majority`` ratio instead of ``majority:minority`` #2077 - Fixed bug where two-way partial dependence plots with categorical variables were not working correctly #2117 - Fixed bug where ``hyperparameters`` were not displaying properly for pipelines with a list ``component_graph`` and duplicate components #2133 - Fixed bug where ``pipeline_parameters`` argument in ``AutoMLSearch`` was not applied to pipelines passed in as ``allowed_pipelines`` #2133 - Fixed bug where ``AutoMLSearch`` was not applying custom hyperparameters to pipelines with a list ``component_graph`` and duplicate components #2133 - Removed ``hyperparameter_ranges`` from Undersampler and renamed ``balanced_ratio`` to ``sampling_ratio`` for samplers #2113 - Renamed ``TARGET_BINARY_NOT_TWO_EXAMPLES_PER_CLASS`` data check message code to ``TARGET_MULTICLASS_NOT_TWO_EXAMPLES_PER_CLASS`` #2126 - Modified one-way partial dependence plots of categorical features to display data with a bar plot #2117 - Renamed ``score`` column for ``automl.rankings`` as ``mean_cv_score`` #2135 - Fixed ``conf.py`` file #2112 - Added a sentence to the automl user guide stating that our support for time series problems is still in beta. #2118 - Fixed documentation demos #2139 - Update test badge in README to use GitHub Actions #2150 - Fixed ``test_describe_pipeline`` for ``pandas`` ``v1.2.4`` #2129 - Added a GitHub Action for building the conda package #1870 #2148 .. warning:: - Renamed ``balanced_ratio`` to ``sampling_ratio`` for the ``BalancedClassificationDataCVSplit``, ``BalancedClassificationDataTVSplit``, ``BalancedClassficationSampler``, and Undersampler #2113 - Deleted the "errors" key from automl results #1975 - Deleted the ``raise_and_save_error_callback`` and the ``log_and_save_error_callback`` #1975 - Fixed ``BalancedClassificationDataCVSplit``, ``BalancedClassificationDataTVSplit``, and ``BalancedClassificationSampler`` to use minority:majority ratio instead of majority:minority #2077
Fixes test for latest dependency bot. Apparently each time the PR is created, it wipes any other changes so I needed to open my own PR 😬
Pandas' update to 1.2.4 broke test_describe_pipeline tests. I believe this is because of the way that pandas treats object columns with float precision via https://pandas.pydata.org/pandas-docs/stable/whatsnew/v1.2.4.html. We call logger.info on the pandas DataFrame, which calls to_string. I've updated this PR to update our test to conform to pandas 1.2.4 since this is simply a difference in printing/logging, but it's worth updating test_describe_pipeline to be tolerant to such changes with pandas behavior in the future (could look into setting some pandas options hardcoded on our end). Unfortunately, this doesn't work here since the float config has also changed here :d