Prostate-Cancer-Dream-Challenge

Team: Data Wizard

**** For Challenge 1a ****

Environment : Linux、Windows 7 Language : Python( Version 2.7.10 )，R studio Packages need to be installed: → Python : pandas、sklearn、Orange → R : survival , timeROC , ROCR , survivalMPL , plyr , caret , discretization , FSelector

Steps:

1. Please change paths of the input files of code if need.
2. Run the “preprocess.py. ” to preprocess the datasets → $python preprocess.py Input : CoreTable_training.csv ,CoreTable_leaderboard.csv

Output : acs.csv, gelg.csv , ven.csv, and CoreTable_leaderboard_new.csv 3. Run the “Final.R” to bin some features.

Input : acs.csv, gelg.csv , ven.csv, and CoreTable_leaderboard_new.csv Output : Train1.csv, Train2.csv, Train3.csv, Test1.csv, Test2.csv, Test3.csv 4. Run the “featureq1a.py” → $python featureq1a.py Input : Train1.csv, Train2.csv, Train3.csv, Test1.csv, Test2.csv, Test3.csv Output : Train_after1.csv, Train2_ after.csv, Train3_ after.csv, Test1_ after.csv, Test2_ after.csv, Test3_ after.csv

5. Run the “Cmodel.R” to fit cox_mpl model

Input : Train_after1.csv, Train2_ after.csv, Train3_ after.csv, Test1_ after.csv, Test2_ after.csv, Test3_ after.csv Output : Q1A.csv

---

**** For Challenge 1b ****

Environment : Linux、Windows 7 Language : Python( Version 2.7.10 )，R studio Packages need to be installed: → Python : pandas、sklearn、Orange → R : survival , timeROC , ROCR , survivalMPL , plyr , caret , discretization , FSelector

Steps:

1. Please change paths of the input files of code if need.
2. Run the “preprocess.py. ” to preprocess the datasets → $python preprocess.py Input : CoreTable_training.csv ,CoreTable_leaderboard.csv

Output : acs.csv, gelg.csv , ven.csv, and CoreTable_leaderboard_new.csv 3. Run the “Q1b.R” to bin some features. → $ Rscript Q1b.R Input : acs.csv, gelg.csv , ven.csv, and CoreTable_leaderboard_new.csv Output : test_Q1b.csv, training_Q1b.csv 4. Run the “final_read_index_q1b.py” → $python final_read_index_q1b.py Input : test_Q1b.csv, training_Q1b.csv Output: result.

---

**** For Challenge 2a ****

Environment : Linux、Windows 7 Language : Python( Version 2.7.10 )，R studio Packages need to be installed: → Python : pandas、sklearn、Orange → R : survival , timeROC , ROCR , survivalMPL , plyr , caret , discretization , FSelector

Steps:

1. Please change paths of the input files of code if need.
2. Run the “preprocess.py. ” to preprocess the datasets → $python preprocess.py Input : CoreTable_training.csv ,CoreTable_leaderboard.csv

Output : acs.csv, gelg.csv , ven.csv, and CoreTable_leaderboard_new.csv 3. Run the “Final.R” to bin some features. Input : acs.csv, gelg.csv , ven.csv, and CoreTable_leaderboard_new.csv Output : TrainAll.csv, TestAll.csv 4. Run the “featureq2.py” → $python featureq2.py Input : TrainAll.csv, TestAll.csv Output : Train_afterAll.csv, Test_ afterAll.csv

5. Run the “final_q2.py” → $python final_q2.py Input : Train_afterAll.csv, Test_ afterAll.csv Output: result