Replace `left`, `right` with `ref` in Unit Tests

[seanlaw]

Currently, when we perform unit test assertions we do something like:

left = naive.stump(T, m)
right = stumpy.stump(T, m)

np.assert_almost_equal(left, right)

However,left and right are also overloaded since we have left and right matrix profiles and indices. It has been proposed that we replace left with ref instead and either replace right or just name it appropriately (maybe comp for "computed"/"comparison"):

ref_mp = naive.stump(T, m)
mp = stumpy.stump(T, m)

np.assert_almost_equal(ref_mp, mp)

or

ref_mp = naive.stump(T, m)
comp_mp = stumpy.stump(T, m)

np.assert_almost_equal(ref_mp, comp_mp)

[sergiobuj]

Hello @seanlaw, I would like to help with this issue.

Just by looking briefly at the project, I see 37 files that need to be updated. Here's the list:

I think the ref_mp/comp_mp reads better than ref_mp/mp.

How would you prefer I opened the pull requests? I'm not too fond of big PRs, but I can accommodate based on the maintainer's preferences.

I was able to install and run tests fine, so I think I'm ready to start (and the documentation was fine.)

PS: I was looking for issues tagged with "good first issue", and that's how I got to the project.

[seanlaw]

@sergiobuj Thank you for your interest in contributing to our project! We can certainly use the help and I am here to assist.

Just by looking briefly at the project, I see 37 files that need to be updated. Here's the list:

In addition to the 37 files listed above, I think you are missing 3 additional tests (40 in total):

tests/mstumped.py
tests/test_mstumped_one_subsequence_inf_self_join.py
tests/test_mstumped_one_subsequence_nan_self_join.py

I think the ref_mp/comp_mp reads better than ref_mp/mp.

Perfect! I agree with your proposal so please go ahead with ref_mp/comp_mp

How would you prefer I opened the pull requests? I'm not too fond of big PRs, but I can accommodate based on the maintainer's preferences.

I really appreciate your consideration! I've taken a look at whether or not it makes sense but I think that this will be a small enough change (even though it is touching 40 files) so I'm not too concerned as long as the tests pass. However, I wouldn't be opposed to it if you wanted to submit, say, 8 separate PRs that updated 5 test files each and it may make it easier to review. Again, this is only a suggestion and I trust your judgement.

I was able to install and run tests fine, so I think I'm ready to start (and the documentation was fine.)

Wonderful! Please proceed and know that I am here to help. Thank you again.

[seanlaw]

@sergiobuj Just a heads up that I made some minor changes to some of the tests in PR #261 . It shouldn't really affect what you're doing but it may result in some merge conflicts that I can resolve. However, if you haven't gotten started yet then I recommend pulling the newest commit and starting there.

[sergiobuj]

@seanlaw thanks for the heads up. I did start with the changes. I will rebase and let you know if I have issues with any conflicts.

[seanlaw]

To save you time from trying to figure out what changed, in the test_gpu_aamp.py and test_gpu_stump.py, I simply swapped the positions of the variables T_A and T_B in several test functions.