We provide the source code for GDOT on synthetic, semi-supervised and supervised experiments conducted in the main paper.

In synthetic folder, we provide data used for OOD generalization (Section 5.1), domain adaptation on CSBM(Section 5.2) and on other two synthetic graphs.
    1. OOD generalization, run
        python synthetic/main_synthetic_1.py --bias_type="hybrid" or "structure"
    2. Domain adapatation on CSBM, run
        python synthetic/main_synthetic_2.py --bias_type="hybrid" or "structure" 
    3. Domain adaptation on syn-cora and syn-products
        python synthetic/main_synthetic_3.py --method=gjdot --dataset="syn-cora" or "syn-products"
On semi-supervised learning and supervised learning, we upload the source code in semi-supervised/ and supervised/. 
Due to the space limit, we are not able to upload the full data used in these two experiments. We will make them public upon acceptance.
