Yoshiki Vázquez Baeza
Yoshiki Vázquez Baeza
Currently an error will be raised when the fingerprint table is empty. However this can occur by one of two reasons. The first being that no fingerprints are predicted and...
If the variance is zero for a PCoA dimension, then the returned features matrix will have those dimensions as NaN. This can be fixed by either filtering zero-variance dimensions or...
The following properties are not documented `name`, `length`, `parent`, `children` and `id`. Currently they render like [this](http://scikit-bio.org/docs/0.5.4/generated/skbio.tree.TreeNode.html#skbio.tree.TreeNode): For the update to napoleon we removed that section (because they rendered awkwardly)....
Only visible when the `Fit bar widths to the plot's default width?` checkbox is not checked.
The input tree should be validated and raise an appropriate exception in these cases: - [ ] duplicated tip ids are present. ~~- [ ] unrooted trees are passed.~~ ~~-...
- [ ] Check there are no negative count values. - [ ] Check there are at least 2 samples.
Specifically this test from skbio's test suite fails: ```python def test_weighted_minimal_trees(self): # two tips tree = TreeNode.read(StringIO('(OTU1:0.25, OTU2:0.25)root;')) actual = self.weighted_unifrac([1, 0], [0, 0], ['OTU1', 'OTU2'], tree) expected = 0.25...
This means that ssu doesn't support these calculations, so a slower implementation has to be used.
There's a few cases where this happens: - The counts table is all full of zeros (using weighted normalized, weighted unnormalized and unweighted unifrac). - The counts table is empty...
In #18 and maybe elswhere there's a few patterns that could be abstracted into a few macros, for example ```c if(obs_indices_resident == NULL) { fprintf(stderr, "Failed to allocate %zd bytes;...