Excuse me. I tried the dataset, e.g., PubMed. from data
| # of nodes : 19717
| # of edges : 44325.5
| # of features : 500
| # of clases : 3
| # of train set : 60
| # of val set : 500
| # of test set : 1000
It reports the edge number is 44,325.5. However, many papers point out that the edge number of PubMed is 88,648.
Why does it only have half size?
Excuse me. I tried the dataset, e.g., PubMed. from data
It reports the edge number is 44,325.5. However, many papers point out that the edge number of PubMed is 88,648.
Why does it only have half size?