Meng Qu

Thank you for pointing out the bug in the codes. Could you provide some detailed information about the error, or provide the dataset you are using? Thanks.

请问这个问题是在哪里遇到的呢？

Hi Zijing, Thanks for the question. Our codes are not compatible with the latest torch-geometric package (2.0.X). You may consider using an older version (1.X.X) instead.

Sorry that we missed the message. Just as wead-hsu pointed out, you can `cd cppext` and run `python setup.py install`. You could also find the commands at: https://github.com/DeepGraphLearning/RNNLogic/blob/318d1f598b615a7697be0cbdc545f1f575eaf167/codes/run.py#L35 Thanks, Meng

Thanks for your interests in RNNLogic! I am wondering what kinds of data do you need? the trained predictor, trained generator, or the generated logic rules?

Thanks for your interest, and very sorry for the late response. We have refactored the codes, and the new codes are in the folder RNNLogic+, which are more readable and...

Thanks for your interests! 1. For psi_s(y_s), we actually tried both options, i.e., (1) directly using `pred_node` or (2) using `sum_s psi_{st}(y_s,y_t)`. These two options yielded close results, and we...

Thanks Wead! I have forwarded the issue to Junkun, the co-first author who implemented rnnlogic w emb. He will figure it out. Meng

> It seems due to the default hyperparameters after inspecting the log. The predictor_num_epoch is so small that the learning is not adequate. This is indeed a problem. The codes...

Thanks for your interests, and this is a good question. The reason is that z_I here is a set of logic rules, so the prior and posterior are defined on...