Reduce use_covariates threshold

[eccabay]

Moves to defaulting to use_covariates=False in more cases, since it produces significant slowdown without much benefit for larger datasets

[codecov]

Codecov Report

Merging #3868 (332e3ba) into main (7438c15) will increase coverage by 0.1%.
The diff coverage is 100.0%.

@@           Coverage Diff           @@
##            main   #3868     +/-   ##
=======================================
+ Coverage   99.7%   99.7%   +0.1%     
=======================================
  Files        344     344             
  Lines      36185   36191      +6     
=======================================
+ Hits       36048   36054      +6     
  Misses       137     137             

Impacted Files	Coverage Δ
evalml/automl/automl_search.py	99.5% <ø> (ø)
...omponents/estimators/regressors/arima_regressor.py	100.0% <100.0%> (ø)
evalml/tests/automl_tests/test_automl.py	99.5% <100.0%> (+0.1%)	⬆️

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

[jeremyliweishih]

LGTM just a suggestion

[christopherbunn]

LGTM, just a question

[christopherbunn]

Not a blocker but is there a reason why this is 7 and not like 5 or 10? Just very curious haha

[eccabay]

Nope, no strong motivation behind it. It was just a bit of a bump down, and without a more comprehensive set of datasets there was no way to test where the best line in the sand would be. Since these results seemed ok, I just stuck with it.