Leighton Pritchard

I thought I'd fixed that error on `master` - do you have a minimal reproducible example as a script and set of input files I can try? You can upload...

We've established that ANIm - as originally implemented - is not a true metric. But we still have to identify a method that does constitute an actual metric (though it...

Closed due to lack of activity.

Fixed with [1f10a6c](https://github.com/widdowquinn/pyani/commit/1f10a6cd78e2d1dcee77f1cc4451cf5ea563841f).

Hi Dan, I'd forgotten that I'd written this. You're right that it isn't complete - I should probably finish it off. My apologies, L.

Hi @ZiliaMR - if your NUCmer run is failing there are two likely possibilities: 1. One of the pairwise genome comparisons is between two organisms that are so dissimilar NUCmer...

Looking through the changes, I see that the calculation looks sensible, and the changes to reporting etc. seem OK - good job! There are a few snags, some of which...

We're getting quite a bit discrepancy in the concordance test for ANIm: ```text > assert result_pid - tgt_pid == pytest.approx(0, abs=tolerance_anim) E assert array([[0.0, ... dtype=object) == 0 ± 1.0e-01...

I've modified the expected output in `test_anim_delta()` so that tests pass, assuming that the calculations are correct. We should go through the calculation for that test file manually and confirm...

> However, we later agreed that the best option would be to calculate the average across all alignments. This would be calculated as follows (assuming that there are two alignments...