RGCN in pytorch

[jdiaz4302]

I see that the current repo has the RGCN.py file from river-dl with the model code written in tensorflow. This PR contains code for 2 versions of the RGCN (v0 = river-dl implementation, v1 = paper equations) written in pytorch. This is just for sharing the model code, giving an overview of it, and maybe doing some further changes.

[jdiaz4302]

Here is some minimal code for using the models

import torch
import numpy as np
from river_dl.RGCN_v1 import *

data = torch.rand([455, 365, 16])
A = np.random.normal(size = [455, 455]) 

model = RGCN_v1(input_dim = 16,
                hidden_dim = 20,
                adj_matrix = A,
                recur_dropout = 0, 
                dropout = 0)

out, (h, c) = model(data)

def count_parameters(model):
    return sum(p.numel() for p in model.parameters() if p.requires_grad)

count_parameters(model)

RGCN_v1 and RGCN_v0 can be used interchangeably. When switching between the two, you ~should~ will use less parameters with RGCN_v1 (which will be displayed with the last line of code above). For that example, RGCN_v0 has 5461 parameters while v1 has 3401 which is a pretty huge % change.

The example uses random data of size [455, 365, 16] to represent one batch from the river-dl data associated with the [455, 455] adjacency matrix.

I have these models coded to output the sequence of predictions along with the last h and c state. This can be easily modified to output the list of h and c states if that is preferred. Also, we can just output the sequence of predictions if that is more compatible with the GWN and you're not interested in states or DA.

[SimonTopp]

@jdiaz4302, just saw this PR after submitting my updates that include the RGCN and RGCN_v1 models. If you'd rather incorporate them here let me know and I can remove them from my PR.

[jdiaz4302]

My intention was opening a PR from SimonTopp:RGCN_torch to SimonTopp:main but apparently I messed that up 😅 . I'll close this out and copy-paste my stuff into that PR