peterlipan/asd_gp_gcn

ASD diagnosis on ABIDE I

asd_gp_gcn's Issues

Accuracy problem

Great job! I used the commands

python download_ABIDE.py
python main.py
But I got the results:

GCN 10 fold test set results, loss = 0.691561, accuracy = 0.517241
GCN 10 fold val set results, loss = 0.632226, accuracy = 0.576923
Mean Accuracy: 0.522257

I haven't changed any parameters, and I would like to ask you if there might be other possible issues that are preventing me from replicating your results.

Graph pooling and LR

Great job！I have a little problem about the Graph pooling and LR. Does this LR classifier directly classifies downsample datas or deal with the features after GCN model? Could you share you source code about the Graph pooling and LR.I am looking forward to your answer.Thanks!

error

Traceback (most recent call last):
File "/download_ABIDE.py", line 83, in
brain_graph(logs, atlas, os.path.join(args.root, 'ABIDE', 'raw'), path)
File "/construct_graph.py", line 107, in brain_graph
cols = list(pd.read_csv(file_path, sep='\t').columns.values)
UnboundLocalError: local variable 'file_path' referenced before assignment

Accuracy

Hello, I am very interested in your work, but when I reproduce the code. The final result was
"GCN 10 fold test set results, loss = 0.777685, accuracy = 0.494253
GCN 10 fold val set results, Loss = 0.720890, accuracy = 0.589744
Mean Accuracy: 0.500562 ",
the TRAIN_ACC of MLP is 1.
Could you tell me what the problem is? Looking forward to your reply!

TUDataset- HTTP Error 404

Excellent work,

I encountered an issue when running this line of code:
abide_dataset = TUDataset(args.data_dir, name='ABIDE', use_node_attr=True)
It results in an HTTP Error 404: Not Found.

Is this because the URL in the newer version of torch_geometric.datasets's TUDataset is incorrect? How can I resolve this problem?

Additionally, I'm curious to know why we need to download ABIDE.zip after data processing is already completed in construct_graph. What is the purpose of downloading this file?

Thank you very much for your explanation and assistance!