Pavan Ramkumar

Implementing RGCNs

3

Thanks for your really great work! I'd like to know if you plan to extend this GAE implementation for learning embeddings of multi-relational graphs (`N x N x M`) as...

Suggestion by Vincent Ferrera at Columbia: [Paper](http://biorxiv.org/content/early/2016/02/26/041301.full.pdf+html) [Matlab Code](https://github.com/belarius/matlab-ARRIS) Thanks @belarius ! Welcome to submit PR

store bootstrap CIs of params if requested in a returnable called `stats`

- population summary of fitted preferred features (histogram) - population summary of fitted preferred features vs. neuron

[WIP] distribution classes

5

closes #350 TODOs - [x] make it work for all distribution classes - [x] add tests for BaseDistribution class - [x] make tests and examples pass locally - [x] remove...

ensure reproducibility of negative bionomial on toy dataset with R

1

Dataset: https://stats.idre.ucla.edu/r/dae/negative-binomial-regression/ R code: ``` require(foreign) require(ggplot2) require(MASS) dat

Pass distribution objects into `GLM` and `GLMCV` class

1

Motivation: as we expand the scope of models supported, it may be easier for contributors to add new distributions if we refactor the class to accept distribution-specific params and corresponding...