qitianwu / NodeFormer

The official implementation of NeurIPS22 spotlight paper "NodeFormer: A Scalable Graph Structure Learning Transformer for Node Classification"
291 stars 28 forks source link

Accurate scores of Figure 2 in the paper #6

Closed AimeJiayue closed 1 year ago

AimeJiayue commented 1 year ago

Hi, your paper is interesting and I would like to follow-up your work. But I noticed that when it comes to the experiment on cora, citeseer, deezer-europe and actor (Figure2 in the paper) there are only plots but not the accurate scores. Would it be convenient to share the accurate scores used for plotting? Thanks in advance.

qitianwu commented 1 year ago

Hi, the exact scores of Fig. 2 are presented in the result table of our Readme.md on the first page.

See here: https://github.com/qitianwu/NodeFormer#key-results

AimeJiayue commented 1 year ago

Hello, I apologize if I didn't express my question clearly. What I would like to know is the scores of all the models (all the models used to draw the graph, including GCN, dropedge, etc., except the Nodeformer) on those four datasets. This would allow us to better follow-up your work, thanks a lot!

qitianwu commented 1 year ago

OK I see. I have uploaded the script for plotting the figure where the testing scores are contained.

See the plot_main.ipynb in the main branch: https://github.com/qitianwu/NodeFormer/blob/main/plot_main.ipynb

AimeJiayue commented 1 year ago

It was really kind of you, thanks a lot.