readme

1.generatedata.py and generatedata4.py are for generating synthetic data, the worth of player are generated from a uniform distribution.
  generatedata4.py is for 4 players, and generatedata.py is for 5 players
  You can generate one test for each run, the data will be stored in Excel_test.xls, and you should copy them in 16or1.xlsx, test1.xlsx, and 16or2.xlsx,test2.xlsx
2. 16or1.xlsx, test1.xlsx, and 16or2.xlsx,test2.xlsx is the synthetic data generated by step 1
3. We have tested 3 sample methods: Random sampling, Stratified sampling,information-driven sampling
   For 4 players, we have shapleyForR.py,shapleyForS.py,shapleyForI.py for Random sampling, Stratified sampling,information-driven sampling
   For 5 players, we have shapleyForR5.py,shapleyForS5.py,shapleyForI5.py for Random sampling, Stratified sampling,information-driven sampling
4. R1, S1, I1 is the experiment result by 4 players, R51, S51, I51 is the experiment result by 5 player
   You can run splotF3.py to get the  Spearman correlation and MSE between appropriate and real Shapley value with 3 methods.
   And splotFSM.py to get the  Spearman correlation and MSE between appropriate and real Shapley value with each data get by step 3
5. SampleFormaddpg.py for sampling the coalition
6. We record all the results from MPE, and calculate the final result by shapleyFMPE.py
7. Plotf3.py is the approximating Shapley for predator-prey environment with increasing sample size.