Automated Tests Lack Instructions

[jranalli]

The project has tests (though they are failing, see Issue #2 ). But JOSS requests that you describe how to run them.

You may wish to consider implementing Continuous Integration using Github Actions to run these tests automatically with each commit, but that's not required for JOSS.

One additional suggestion is to improve the coverage. Optimization is mostly covered but it would be nice to see the tests include the interpolation, isoconversional methods and kinetic models at least.

[alan-tabore]

I added some tests and will continue adding test (especially for the isoconversional analysis.

I also added GitHub Actions to run the tests automatically. Thanks again for the advice.

[jranalli]

Please see here: #2 (comment)

test_optimization doesn't actually do anything right now. Even if you modify it, I believe it would just be testing for argument agreement rather than the actual optimization functionality. The other tests have decent coverage, but you might wish to run them using the coverage module to see what they actually do cover. Just for reference:

test_interpolation covers everything but the two cubic spline functions.
test_kinetic_models covers all but compute_extent_and_rate, hess_for_rate_for_kamal and jac_for_rate_for_kamal
'test_optimization` like I said would only cover the function definitions.

I'm not here to nitpick test coverage at a detailed level, that's just suggestion. But I would say the test that doesn't do fails to find any functions to test should be fixed to be consistent with the journal's expectations that testing is performed.

[alan-tabore]

Thanks again for your feedback @jranalli .

I've corrected the test_optimization module to really test the arguments.
I've also added tests to actually check the correct implementation of the cost functions (or rss functions).

Would you also recommended to test the validy of the Scipy optimization methods ?
It would be really tricky to do since I'm just using those function so I did not create a separated module and directly used the optimization algorithms in the main.py (file handling all GUI operations and user inputs)

I'll remove the two cubic spline functions in the optimization module as well as the hess_for_rate_for_kamal and jac_for_rate_for_kamal function since their use is limited and prone to errors.

I still need to cover the compute_extent_and_rate function.
I'll also check the coverage (thanks for the tip, I didn't know the coverage package)

I'll let this issue open until I've completed these last steps.

[jranalli]

Would you also recommended to test the validy of the Scipy optimization methods ? It would be really tricky to do since I'm just using those function so I did not create a separated module and directly used the optimization algorithms in the main.py (file handling all GUI operations and user inputs)

I don't think that's absolutely necessary. It would be up to you though. If you want to be able to easily identify problems with your code that arise due to changes in dependencies, testing is necessary. Again though, that's part of what makes review for a GUI app a bit different than what I've done otherwise for JOSS.