Check test coverage

[billbrod]

Probably worth using something to check how complete our test coverage is (that is, are we not missing a test): see here for general discussion and pytest-cov for a library we could use.

[billbrod]

From @rodrigocam:

Code coverage is a polemic field in software engineering because you can have a high code coverage (80-90%) and still not test the core functionalities of your code, but it still a good software measurement associated with other metrics. Code coverage can be extracted locally (in python with https://pypi.org/project/coverage/) and also can be extracted with an automated tool. I recommend codecov (python example here https://github.com/codecov/example-python), it haves a nice integration with Travis CI that you already use. Basically what it will do is every time Travis CI runs (pull request merged or commit) it will run test coverage and updates the data in codecov (also updating your badge). coveralls it is also a good tool that does the same!

Software metrics are very important aspects to guide decisions, if you want more metrics codeclimate is a good tool, it gives you a grade to your code based on a few metrics and spots the parts of your code that can be improved (code duplication and code too complex).

[billbrod]

Looks like all of those require public repos to be free, so look into after we make them public.