Add test_size parameter to ClassImbalanceDataCheck

[freddyaboulton]

Pull Request Description

Fixes #3334

---

After creating the pull request: in order to pass the release_notes_updated check you will need to update the "Future Release" section of docs/source/release_notes.rst to include this pull request by adding :pr:123.

[codecov]

Codecov Report

Merging #3341 (370104d) into main (f4a671d) will decrease coverage by 0.7%.
The diff coverage is 100.0%.

@@           Coverage Diff           @@
##            main   #3341     +/-   ##
=======================================
- Coverage   99.6%   99.0%   -0.6%     
=======================================
  Files        329     329             
  Lines      31977   32000     +23     
=======================================
- Hits       31847   31649    -198     
- Misses       130     351    +221     

Impacted Files	Coverage Δ
evalml/data_checks/class_imbalance_data_check.py	100.0% <100.0%> (ø)
...ta_checks_tests/test_class_imbalance_data_check.py	100.0% <100.0%> (ø)
evalml/automl/pipeline_search_plots.py	17.9% <0.0%> (-82.1%)	⬇️
...l/tests/automl_tests/test_pipeline_search_plots.py	23.3% <0.0%> (-76.7%)	⬇️
...ests/automl_tests/test_automl_search_regression.py	74.3% <0.0%> (-20.4%)	⬇️
.../automl_tests/test_automl_search_classification.py	83.8% <0.0%> (-12.6%)	⬇️
evalml/tests/automl_tests/test_automl_utils.py	91.1% <0.0%> (-8.9%)	⬇️
...lml/tests/automl_tests/test_iterative_algorithm.py	92.3% <0.0%> (-7.7%)	⬇️
evalml/automl/utils.py	98.5% <0.0%> (-1.5%)	⬇️
evalml/automl/automl_search.py	99.6% <0.0%> (-0.1%)	⬇️
... and 1 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update f4a671d...370104d. Read the comment docs.

[chukarsten]

All this increasing of the targets make me wonder if there's any consolidation that can be done in the evalml/tests/data_checks_tests/test_class_imbalance_data_check.py module to perhaps push some of those synthetic targets out of the individual tests and either to the module level or to a conftest.py further up. But other than that, looks gucci.

[chukarsten]

Hyper nit - won't you need a Raises: block for the new ValueError? I guess not given the existing ValueError isn't complained about.

[angela97lin]

+1 to ValueError, lame that it's not caught by our linter because we move our docstring from init to our class docstring (IIRC for docs) so it doesn't automatically pick it up 😢

[freddyaboulton]

My guess is that darglint doesn't check for raises blocks in constructors. @angela97lin do you know?

Lol Angela and I commented at the same time. Adding the missing Raises block!

[angela97lin]

Yeah @freddyaboulton, I don't think it does a good job for class docstrings in general (I think I recently came across a case where we added a new parameter but it didn't complain about the missing docstring).

If we had this docstring in the constructor (init) it might actually pick it up, but IIRC, then our docs won't look like this: https://evalml.alteryx.com/en/stable/autoapi/evalml/data_checks/class_imbalance_data_check/index.html#evalml.data_checks.class_imbalance_data_check.ClassImbalanceDataCheck

We'd have to add back init and click on the init method link specifically to see the parameters, rather than just at the class page.

[angela97lin]

Clever use of numpy array multiplication to update the tests, LGTM! Thanks @freddyaboulton 😁

[angela97lin]

+1 to ValueError, lame that it's not caught by our linter because we move our docstring from init to our class docstring (IIRC for docs) so it doesn't automatically pick it up 😢