search

dieterich-lab / rp-bp

Rp-Bp is a Bayesian approach to predict, at base-pair resolution, ribosome occupancy and translation.
MIT License
7 stars 5 forks source link

Add test to compare expected output to actual output for example #76

Closed bmmalone closed 6 years ago

bmmalone commented 7 years ago

The example already includes an expected-orf-predctions which gives the output expected when running Rp-Bp on the C. elegans example.

However, it is difficult to compare the actual output to the expected output. Additionally, if we eventually want to add continuous integration to this project, we need to have the tests work through nose (or unittest).

Thus, we should add a test which makes this comparison and is "nose-friendly".

bmmalone commented 7 years ago

The documentation should also be updated to reflect the new approach to testing.

eboileau commented 6 years ago

I agree with nose test integration in the longer term, though the very essence of unit testing seems a bit incompatible with the example test, as we would most likely want to keep the files and results after running Rp-Bp on the C. elegans test data. Whether we should provide a template script that runs everything or only a detailed work flow as in running-example, I don't know. I would rather keep things as they are now, and clarify/update the docs (and the data/configs) where necessary. As for proper benchmarking, we can propose a number of options (e.g. bedtools). One last comment: using the latest Rp-Bp release (Dec.17), the actual output does not compare exactly with the expected output...

eboileau commented 6 years ago

The dataset has been moved over to the new cloud. All docs regarding the example updated, I will add the new links. My opinion is that there is enough information to run the example, so I will close this for now.