prevent ImportError due to scikit-optimize and sklearn incompatibility

[basnijholt]

Description

All tests fail with

______________ ERROR collecting adaptive/tests/test_learnernd.py ______________
ImportError while importing test module 'D:\a\1\s\adaptive\tests\test_learnernd.py'.
Hint: make sure your test modules/packages have valid Python names.
Traceback:
.tox\py36-alldeps\lib\site-packages\_pytest\python.py:511: in _importtestmodule
    mod = self.fspath.pyimport(ensuresyspath=importmode)
.tox\py36-alldeps\lib\site-packages\py\_path\local.py:701: in pyimport
    __import__(modname)
.tox\py36-alldeps\lib\site-packages\_pytest\assertion\rewrite.py:152: in exec_module
    exec(co, module.__dict__)
adaptive\tests\test_learnernd.py:7: in <module>
    from .test_learners import generate_random_parametrization, ring_of_fire
.tox\py36-alldeps\lib\site-packages\_pytest\assertion\rewrite.py:152: in exec_module
    exec(co, module.__dict__)
adaptive\tests\test_learners.py:31: in <module>
    from adaptive.learner.skopt_learner import SKOptLearner
adaptive\learner\skopt_learner.py:4: in <module>
    from skopt import Optimizer
.tox\py36-alldeps\lib\site-packages\skopt\__init__.py:54: in <module>
    from .searchcv import BayesSearchCV
.tox\py36-alldeps\lib\site-packages\skopt\searchcv.py:16: in <module>
    from sklearn.utils.fixes import MaskedArray
E   ImportError: cannot import name 'MaskedArray'

See scikit-optimize/scikit-optimize#902

Checklist

• Fixed style issues using pre-commit run --all (first install using pip install pre-commit)
• pytest passed

Type of change

Check relevant option(s).

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

[akhmerov]

Please open an issue to track the regression, and add a TODO comment in the code with the explanation why we catch import error and referring to the issue.

Also: did you find this problem reported upstream? If yes, please link the upstream bug report, otherwise please report it.

[codecov-commenter]

Codecov Report

Merging #278 into master will increase coverage by 0.18%.
The diff coverage is 83.33%.

@@            Coverage Diff             @@
##           master     #278      +/-   ##
==========================================
+ Coverage   80.49%   80.67%   +0.18%     
==========================================
  Files          34       34              
  Lines        4588     4591       +3     
  Branches      828      828              
==========================================
+ Hits         3693     3704      +11     
+ Misses        768      764       -4     
+ Partials      127      123       -4     

Impacted Files	Coverage Δ
adaptive/learner/learnerND.py	59.72% <ø> (ø)
adaptive/notebook_integration.py	18.32% <ø> (ø)
adaptive/tests/test_learners.py	98.70% <0.00%> (ø)
adaptive/tests/test_runner.py	94.93% <ø> (+2.25%)	⬆️
adaptive/runner.py	70.56% <66.66%> (+1.09%)	⬆️
adaptive/learner/average_learner.py	86.20% <83.33%> (+2.48%)	⬆️
adaptive/tests/test_average_learner.py	94.44% <100.00%> (+1.58%)	⬆️

---

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 1c60897...b71c6e8. Read the comment docs.

[basnijholt]

I've already linked the upstream issue and I will add a todo in the code referring to this PR.