Follow instructions in  GraphSAGE-master/real_data/Readme_Download.txt

Download pre_trained GCN model : https://drive.google.com/drive/folders/1UcQ2PJqHo4QLR5pojz81KXeqwpTlWHYA?usp=sharing
Put it in GraphSAGE-master after extracting


Train:
sh train_script.sh

Test
sh test_script.sh




Summary:


To train SUP gsage

Run
 cd GraphSAGE-master
 python train_multiple.py

Run
python3 predicte_multiple_for_train.py


To train RL:
cd..
python train_RL.py



Testing

predict gsage embeddings

cd GraphSAGE-master

python3 predict_multiple.py



Test RL:
cd ..
python easy_testing.py



To evaluate spread:

Run python im_eval_spread_for_tv.py


#10 simulation graphs are uploaded (GraphSAGE-master/real_data/youtube/TV/test/large_graph/mc_sim_graphs/)
# You can create more simulation graphs for calculating MC using
python spread_pre_process.py

Change 10000 to 10 to run currently without creating more simulation graphs.




To evaluate for more datasets:
Create folder : GraphSAGE-master/real_data/DATASET_NAME/
Put edge file edges.txt in GraphSAGE-master/real_data/DATASET_NAME/

run python convto_nx.py DATASET_NAME


For example stack
python convto_nx.py stack


Download more datasets using
cd GraphSAGE-master/real_data
sh download_datasets.sh






Below steps are in case you wish to generate training data and train interpolator :
Default is youtube.



Pre-requisistes:
First you need to build IMM ( Since we have used IMM for generating training labelled data since its relatively faster.)

cd imm_baseline/im_benchmarking-master/sidm029_im_benchmark/Codes/IMM/
make


Go to home folder of the project 
cd ../../../../../../


To get labelled training data: 
We have used IMM for generating training labelled data since its relatively faster.

Run below to get labels for training dataset
(Default is youtube)
 
./get_train_labels_single.sh



For training  interpolator :

Run to get labels for training dataset for small size subgraphs of train data
Default is youtube
 
 Run 
./get_train_labels_size_var.sh


To get interpolator weights

cd GraphSAGE-master
python3 size_Var_rank_analysis_getlowest.py