Ben Pedigo

Refactored graph matching code and added many new features

4

- [x] Does this PR have a descriptive title that could go in our release notes? - [ ] Does this PR add any new dependencies? - No - [x]...

Order of arguments in docs sidebar

1

I suggest the sidebar docs should go: - Install - Tutorials - Reference (maybe this could be swapped with tutorial?) - Everything else discussed this with @daxpryce - the way...

Align module enhancements

1

I have code I need to PR which adds: - [ ] the right way to do alignments for embeddings of directed networks - [ ] the ability to use...

Networkplot should be able to handle numeric rather than categorical colormapping

2

Currently, the color mapping that happens in networkplot assumes that the variable is categorical, more or less (see here https://github.com/microsoft/graspologic/blob/690de9b6284441ccaab1d79f08b3718c93d28072/graspologic/plot/plot.py#L1356) I am used to the seaborn behavior where numeric variables...

Add 2-sample tests under ER, SBM, density adjusted SBM models

1

I have been developing 2-sample tests under an ER or SBM assumption here: https://github.com/neurodata/bilateral-connectome Considering adding these as functions to the `inference` submodule. Would be more inclined to do so...

Propose removing I-DAD as a laplacian

1

related to #938 I propose that we don't need the I - DAD as an accepted laplacian, because: pros: - our library always does SVD using top singular values, when...

On this line https://github.com/microsoft/graspologic/blob/49d6929f14b57d74ba80f903a7aad19f5e4dafa2/graspologic/match/qap.py#L446 perm_S is referring to the unseeded vertices of `B`, when it should really be the unseeded vertices of `A`: `S22 = s_value[nonseed_A, n_seeds:]`

Graph matching code structure discussion

3

## Is your feature request related to a problem? Please describe. - Graph matching code is great as is if just following FAQ/SGM exactly - I'd like to generalize the...

[BUG] "Parameters" showing up twice in docs for class methods

1

Somehow changes in #907 made it so that "parameters" shows up twice, once from the type hints and once from what we wrote in the doc string. Doesn't feel like...