Sergio Rey

@sjsrey @knaaptime @ljwolf @martinfleis agreed to [rename W to Graph and move the namespace](https://github.com/pysal/libpysal/commit/ded062584a4d3d9c4ed1c668119a6e3b918aead6).

@Palak-B can you let us know what version of numba you are running as this is passing on the [ci](https://github.com/pysal/esda/actions/runs/5427360269/job/14699410803#step:7:61)?

xref https://github.com/pysal/libpysal/issues/526

xref https://github.com/pysal/libpysal/issues/522

Building on an environment with 3.10 locally works fine. I think we have the CI setup to do the same, so perhaps we close this?

@JosiahParry the determination of the pseudo-values is done with the following: https://github.com/pysal/esda/blob/7f831081c987e0d4bcf810c78b80519569694e9f/esda/crand.py#L222-L224 which is for the local statistics. This follows the logic we use for a global statistic (which is...

@ljwolf the outside counting method we have is a one-tailed procedure that follows from logic in geoda. It does the selection of the tail for the user. I think we...

I see what you are getting at, and I agree with the points you make. My interpretation is the logic for the permutation based inference was born in the "pre-local"...

I agree on the new logic. I'm a bit hesitant to deprecate the current behavior as it is based on what [geoda does](https://github.com/GeoDaCenter/geoda/blob/0465201357e1157975e663eb3b8a707648d3ea8b/Algorithms/lisa_kernel.cl#L197). Might we consider a `policy="hope|geoda"` parameter with...

Look into automated testing of notebooks following [this logic](https://github.com/sjsrey/gds_scipy16/tree/master/content/infrastructure) and [nbconvert approach](https://gist.github.com/lheagy/f216db7220713329eb3fc1c2cd3c7826)