To run HamGNN model

Node classification on CORA 
python train_all.py --dataset cora --task nc --use-att 0 --num-layers 3 --dim 64 --odemap h1extend --patience 200 --epochs 2000

Node classification on CITESEER 
python train_all.py --dataset citeseer --task nc --use-att 0  --num-layers 3 --dim 64 --odemap h1extend --patience 200 --epochs 2000

Node classification on Pubmed 
python train_all.py --dataset pubmed --task nc --use-att 0  --num-layers 3 --dim 64 --odemap h1extend --patience 200 --epochs 2000

Node classification on Disease_nc 
python train_all.py --dataset disease_nc --task nc --use-att 0  --num-layers 3 --dim 64 --odemap h1extend --patience 1000 --epochs 2000

Node classification on Airport
python train_all.py --dataset airport --task nc --use-att 0  --num-layers 3 --dim 64 --odemap h1learn --patience 1000 --epochs 2000

Link prediction results and heterophilic dataset hyperparameter saved in ./hyperparameter folder
The main function is  train_heter.py and tarin_lp_all.py




the function H_net in the paper saved in the ./layers folder 
H_1.py refers to Equation(20)     --odemap h1extend
H_2.py refers to Equation(21)	  --odemap h2extend
H_3.py refers to Equation(52)	  --odemap h3extend
H_4.py refers to Equation(22)	  --odemap h4extend
H_6.py refers to Equation(51)	  --odemap h6extend
H_8.py refers to Equation(23)	  --odemap h8extend
Ode_map.py.Connection_v5extend refers to Equation(19)    --odemap h9extend






