Remove data check for log transformation in make_pipeline

[angela97lin]

Closes #2601. #2679 tracks adding a parameter which would enable users to add preprocessing components to automl search, which users can use to pass the log transformer recommended by an action!

Note that while this may have some scoring implications that may lead to a decrease in performance for datasets with log distributions, this better aligns with the API / flow we have for data check actions.

[codecov]

Codecov Report

Merging #2806 (83c43c9) into main (4027faa) will decrease coverage by 0.1%.
The diff coverage is 100.0%.

@@           Coverage Diff           @@
##            main   #2806     +/-   ##
=======================================
- Coverage   99.8%   99.8%   -0.0%     
=======================================
  Files        297     297             
  Lines      27757   27744     -13     
=======================================
- Hits       27689   27676     -13     
  Misses        68      68             

Impacted Files	Coverage Δ
evalml/pipelines/utils.py	99.2% <100.0%> (-<0.1%)	⬇️
evalml/tests/automl_tests/test_automl.py	99.7% <100.0%> (-<0.1%)	⬇️
evalml/tests/pipeline_tests/test_pipeline_utils.py	100.0% <100.0%> (ø)

---

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 4027faa...83c43c9. Read the comment docs.

[ParthivNaresh]

My beautiful data check 😭 just kidding, this is a good cleanup and data check actions are definitely the way to go if we want to leverage any of these data check outcomes in pipelines in the future.

Also you might need to update the docs here to remove the last part of the second paragraph:

"If you use AutoML.search to try and find the best pipeline, it will automatically run this data check for you and, if a lognormal distribution is detected, will add a LogTransformer to your pipeline to help the model performance!"

[freddyaboulton]

Thank you @angela97lin !

[freddyaboulton]

I think this is technically a breaking change.

[angela97lin]

I always think of a breaking change as something that would cause our API to change, not the implementation, but I'm down to add it so that we're loud about the behavior of AutoML changing 😁

[bchen1116]

LGTM!

[eccabay]

Nice! I really like the solution you found with the warnings test 😄