FIX: transposing ranks for final output. #99
Conversation
Found that standalone cli wasn't consistent with q2. Biases weren't factored into standalone ranks. Now fixed.
|
This is turning into a large pull request. The original intent of this PR was to allow for the ranks to be transposed in order to view more easily in excel and spreadsheet programs. To do this, a new q2-transformer needed to be added to transform ranks to metadata (to view in However, during this refactor, it was discovered that the ranks between q2 and the standalone version weren't completely consistent, namely the biases weren't incorporated into the standalone version. Long story short, the previous iterations of the ranks maybe biased after post-processing. |
Also adding in cool figure, because, why not
|
Since there were some post-processing changes to the outputs, a couple of more checks, namely with the cystic fibrosis dataset and the biocrust soils dataset. Below is the CF dataset, showing that the biplot hasn't changed. Furthermore, the ranks for the microcoleus and pseudomonas rhamnolipids are consistent with what we expect.
|
|
We're going to merge as soon as these tests finish. Then we'll push out a release today. |
Updating documentation in readme accordingly. Addresses #98