Hello! I recently read your paper on RGNN model and am very interested in it. But I had some trouble reproducing your paper.
When I run python train.py directly, I can get results similar to that in the paper.But, when running data_process.py, an error occurs, and the error information is as follows:
I tried to fix this error by modifying lines 95-96 from:
unique_users = users_count.index
unique_items = items_count.index
to：
unique_users = users_count['users_id']
unique_items = items_count['items_id']
However, the result of this is that the accuracy rate is reduced to 0.8106 without any change in the parameters. Here are the results of the run：
Please tell me how to divide the data set correctly to reproduce the experimental results? thank you!
Hello! I recently read your paper on RGNN model and am very interested in it. But I had some trouble reproducing your paper.
When I run python train.py directly, I can get results similar to that in the paper.But, when running data_process.py, an error occurs, and the error information is as follows:
I tried to fix this error by modifying lines 95-96 from:
unique_users = users_count.index
unique_items = items_count.index
to：
unique_users = users_count['users_id']
unique_items = items_count['items_id']
However, the result of this is that the accuracy rate is reduced to 0.8106 without any change in the parameters. Here are the results of the run：
Please tell me how to divide the data set correctly to reproduce the experimental results? thank you!