Remove non-selected-categorical columns from dataframe used to create numeric pipelines in DefaultAlgo

[jeremyliweishih]

If a dataframe contained a categorical column that was not selected by the feature selector, DefaultAlgorithm would send a dataframe with such categorical column to the make pipeline util for the numeric pipeline. This would cause a OHE to be inserted into the numeric pipeline. This change makes it so that a dataframe matching the parameters in the SelectByType is sent to the pipeline util instead.

[codecov]

Codecov Report

Merging #3914 (827f0bc) into main (5e5d652) will increase coverage by 15.6%.
The diff coverage is 100.0%.

@@           Coverage Diff            @@
##            main   #3914      +/-   ##
========================================
+ Coverage   84.1%   99.7%   +15.6%     
========================================
  Files        346     346              
  Lines      36667   36669       +2     
========================================
+ Hits       30816   36536    +5720     
+ Misses      5851     133    -5718     

Impacted Files	Coverage Δ
...valml/automl/automl_algorithm/default_algorithm.py	100.0% <100.0%> (+37.4%)	⬆️
...valml/tests/automl_tests/test_default_algorithm.py	100.0% <100.0%> (+100.0%)	⬆️
evalml/pipelines/utils.py	99.6% <0.0%> (+2.0%)	⬆️
evalml/tests/conftest.py	98.0% <0.0%> (+4.5%)	⬆️
evalml/automl/engine/engine_base.py	100.0% <0.0%> (+5.6%)	⬆️
.../preprocessing/data_splitters/time_series_split.py	96.7% <0.0%> (+13.4%)	⬆️
evalml/automl/automl_algorithm/automl_algorithm.py	100.0% <0.0%> (+16.5%)	⬆️
evalml/automl/utils.py	97.3% <0.0%> (+16.5%)	⬆️
evalml/tuners/tuner.py	100.0% <0.0%> (+24.6%)	⬆️
evalml/automl/automl_search.py	99.6% <0.0%> (+31.0%)	⬆️
... and 26 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[eccabay]

Looks good!

[ParthivNaresh]

Thank you!