WIP: Add tests #36
WIP: Add tests #36
Conversation
|
hm, I'm actually not sure about this yml stuff, I've never used it are these just standard unit tests, or tests in the sense that we test something against data? |
|
@dburov190 I suggest we start with simple easy tests (for a start we just want to ensure that the functions described in the README execute without error; the current method of running the examples will not scale). The tests I've included combines a test for the instantiation of the samplers and transition rate class (implicitly), as well as a test that ensures we can reproduce the results of the random samplers. Ideally we would also validate the algorithms we are using (for example, the implementation of the Kalman filter). We'll see what we have time for. |
| def get_rates(rates_name): | ||
| return getattr(king_county_transition_rates(population=3, random_seed=1234), rates_name) | ||
|
|
||
| transition_rates = king_county_transition_rates(population=3, random_seed=1234) |
There was a problem hiding this comment.
this seems outdated now because scenarios don't have king_count_transition_rates anymore
but I like this test and overall idea, could you massage this into a running test? I'd start adding tests then
There was a problem hiding this comment.
Yes, this PR has gone stale. I agree this PR should be revived.
A low-effort solution would be to implement a few simple tests for our most high-level objects, since this will implicitly test lower-level functionality.
We can also consider running some of the examples during testing.
There was a problem hiding this comment.
I'd actually strip apart some of the examples into shorter tests, but, one way or another, we must start writing them, there are so many bugs right now that it's actually scary :/
The tests are run by navigating to
/testand typingI think we need to add
pytestto theymlas well. @dburov190 and @agarbuno please advise (or simply commit to this PR).Resolves #35