Andrew Docherty

Hi, Thanks for reporting this bug. You are correct that the `DirectedGraphSAGE` layer doesn't currently support the arguments that it claims to – namely `in_samples` and `out_samples`. **This means that...

Hi, Thanks for opening this issue. We would love to have some demos for entity alignment, but currently we do not have the development time to do this. We are...

Hi, Firstly, I'll confess to not understanding entirely what your model is doing. A few comments: * You are adding a very deep network with 9 layers - for GCN...

Hi, I think that you are correct - the Node2Vec embeddings will change between runs even on the same graph. As I understand, the intention in this code is that...

Hi, Thanks for opening this issue. It would be helpful if you could include more information about your problem and if possible some minimal code that replicates the issue. Currently,...

Thanks for the question. Stellargraph does not have any inbuilt plotting capabilty and relies on third party libraries for plotting. The [NetworkX library](https://networkx.org/documentation/latest/tutorial.html#drawing-graphs) includes good capabilities for plotting. You can...

Hi, Thanks for opening an issue. I agree that it would be a great improvement to demonstrate how to make predictions for graph classification problems, as we do for node...

This is a Keras bug. It is pretty subtle and only occurs in the nasty confluence of events that the model has shared layers that are saved in the wrong...

This is a bug fix to Keras model loading that solves this issue: ``` diff --git a/keras/engine/network.py b/keras/engine/network.py index 51ea1d12..5244c2f2 100644 --- a/keras/engine/network.py +++ b/keras/engine/network.py @@ -973,12 +973,10 @@ class...

Added bug to Keras team issues: https://github.com/keras-team/keras/issues/11159