Move sklearn wrapper to separate library

[rwedge]

There's a bug (scikit-learn/scikit-learn/issues/15795) in the 0.22 release of scikit-learn that is breaking one of our test cases. Telling pip to ignore specifically the 0.22 release to avoid this bug.

[codecov-io]

Codecov Report

❗ No coverage uploaded for pull request base (master@cd674b3). Click here to learn what that means.
The diff coverage is 100%.

@@            Coverage Diff            @@
##             master     #835   +/-   ##
=========================================
  Coverage          ?   98.15%           
=========================================
  Files             ?      117           
  Lines             ?    10848           
  Branches          ?        0           
=========================================
  Hits              ?    10648           
  Misses            ?      200           
  Partials          ?        0

Impacted Files	Coverage Δ
featuretools/utils/cli_utils.py	55.31% <ø> (ø)
featuretools/__init__.py	65.51% <ø> (ø)
featuretools/tests/utils_tests/test_cli_utils.py	100% <ø> (ø)
featuretools/version.py	100% <100%> (ø)
featuretools/tests/test_version.py	100% <100%> (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 cd674b3...955609a. Read the comment docs.

[kmax12]

is sklearn only used by the DFSWrapper? if so, maybe we should make it an optional dependency since that functionality is "beta" anyways

[rwedge]

Yeah it's only used in that wrapper

[kmax12]

looks good. just one comment

[kmax12]

i think we can keep this in like we do for the nlp primitives. we just need to mention how to install

https://docs.featuretools.com/en/stable/api_reference.html#natural-language-processing-primitives