Remove unused dependencies

[dsherry]

We list boto3 and boto-core as deps but don't use them in our repo.

s3fs is needed for the docs, but not by our code.

And ipywidgets appears to be needed to run our unit tests and build our docs, but it's not referenced anywhere in our code.

Related to #339

[codecov]

Codecov Report

Merging #451 into master will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master     #451   +/-   ##
=======================================
  Coverage   98.19%   98.19%           
=======================================
  Files         104      104           
  Lines        3260     3260           
=======================================
  Hits         3201     3201           
  Misses         59       59

---

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 fb2ca4f...c59bf41. Read the comment docs.

[angela97lin]

Ah yeah, I think we removed the need for boto3 and boto-core when we moved our dataset offline, nice! :D

[dsherry]

Oooh, I found a hole in our checkin test coverage: we always install dev-requirements.txt before running unit tests. This PR moved ipywidgets to dev-requirements.txt; the checkin tests pass, but then when I create a new virtualenv, install only requirements.txt and try to import evalml, I get an error because ipywidgets is missing:

(evalml_test) ➜  evalml git:(ds_remove_deps) python3
Python 3.7.6 (default, Dec 30 2019, 19:38:28)
[Clang 11.0.0 (clang-1100.0.33.16)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import evalml
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/Users/dylan.sherry/development/evalml/evalml/__init__.py", line 24, in <module>
    from evalml.automl import AutoClassificationSearch, AutoRegressionSearch
  File "/Users/dylan.sherry/development/evalml/evalml/automl/__init__.py", line 7, in <module>
    from .auto_classification_search import AutoClassificationSearch
  File "/Users/dylan.sherry/development/evalml/evalml/automl/auto_classification_search.py", line 4, in <module>
    from .auto_base import AutoBase
  File "/Users/dylan.sherry/development/evalml/evalml/automl/auto_base.py", line 10, in <module>
    from .pipeline_search_plots import PipelineSearchPlots
  File "/Users/dylan.sherry/development/evalml/evalml/automl/pipeline_search_plots.py", line 4, in <module>
    from IPython.display import display
ModuleNotFoundError: No module named 'IPython'

The solution is to a) keep ipywidgets in requirements.txt and b) add test coverage for this (#453 )