Fix _pre_test hanging when generating candidates

[josepablocam]

This PR addresses the issue #1023

We added a time out to the pipeline .fit call.

Running the below on the original codebase should hang, while the patched version should complete as expected.

> cat example.py

import tpot
import numpy as np
import pmlb

X, y = pmlb.fetch_data("Hill_Valley_without_noise", return_X_y=True)

clf = tpot.TPOTClassifier(
  max_time_mins=1,
  max_eval_time_mins=1,
  cv=2,
  config_dict={
      'sklearn.decomposition.SparsePCA':{},
      'sklearn.linear_model.LogisticRegression':{},
  },
  verbosity=3,
)

clf.fit(X, y)

timeout 5m python example.py
echo $?

Note that I chose a timeout of 2 seconds in the patch (an arbitrary choice). I leave the choice of this to the maintainers, who are likely to have a better sense of a reasonable value.

Questions:

• Do the docs need to be updated?
  No

• Does this PR add new (Python) dependencies?
  No

[coveralls]

Coverage increased (+0.008%) to 96.755% when pulling 4bbb622 on josepablocam:proposed-fix into 6efc21e on EpistasisLab:master.

[weixuanfu]

Thank you for the PR and it looks good for fixing the issue #1023. I will merge it to dev branch for now and add this patch to next version of TPOT.