GSRR: Graph Similarity and Resilience Ranking

utils

dataprocess

GraphConstruct

utils

model

GAT

model_cuda

model_cuda2

NodeEmbeddingModule2：modify GAT

MGC-RM

1

perturbation

Generate a perturbation graph

perturbation2

Generate a perturbation graph

packaged into a function

2

MFC_RMF

Multi-Granularity Cross Representation and Matching

for i in range(perturbed_a.shape[0]) Perturbation graph are involved in training

plt.savefig
torch.save pth

MFC_RMF2

1. Specify a graph pair to train and save the model.

2. Add the model overload module to generate the similarity score of the remaining graph pair and save it as a file. (Nearly 4 hours)

3. Save prediction_loss as txt and fig to evaluate the difference in the effect of the graph

np.savetxt('./similarity score/z'+ 'GP_'+str(j)+'_node_'+str(node)+'_data_'+str(data_num)+'model_'+str(formatted_time)+'.txt', z_p)

np.savetxt('./prediction_loss/GraphPair_' + str(j) + '_n' + str(node) + '_d' + str(data_num) + '_Prediction_Loss_epoch_' + str(args.max_epoch) + '_lr_' + str(args.lr) + '_' + str(formatted_time) +'.txt', loss_history2)

MFC_RMF2cuda2

3

plotpredictloss

plotscore

4

PageRank2

weights PageRank

5

Perform

plotpredictloss

plotscore

6

no-readout

Resilience

1

resilience-cpu

cuda Unused

The oldest version

resilience_cuda

CrossEntropy

Basic version

resilience_train_test

model test:

use cuda

with plot

resilience_eval

model test:

use cuda

2

plotscore2

line charts

plotscore-r

for Rg

plotscore-y

for predict data

3

R-Perform

Evaluate metric calculations