Run the program as follows:

python one_run.py {method} {dataset} {multiset/minimal #f threshold} {random seed}

e.g. if you want to test the diabetes dataset using mrmr you call "python one_run_corrected_crossfold.py 6 2 0.96 42"

Call script with the following arguments:



Dataset:

        1: "breast_cancer"
        2: "diabetes"
        3: "diabetes_mrmr"
        4: "medical_cognition"
        5: "medical_frail"
        14: "parkinsons"
        15: "lungcancer"
        18: "lymphography"
        20: "heart_disease"
        21: "sick"
        22: "primary_tumor"
    
Feature selection method:

        1: "none"
        6: "mrmr"
        17: "deltadebugging_new_dd"
        18: "deltadebuggingshuffle_new_dd"
        42: "POSS"
        43: "hho_test"







Example output: 

You call the script as "python one_run_corrected_crossfold.py 18 15 1.0 42"

You will recieve an output as follows: 
---------------------------------------------------------------------
lungcancer_deltadebuggingshuffle_new_dd_False_1
Best accuracy: 0.87143
Feature reduction: 54 --> 6
Total Time taken for tuning ML + FS: 43158.10238 seconds
Number of multisets: 49
Avg Size of multisets: 23.26530612244898
Minimal feature set size: 2
---------------------------------------------------------------------
Best split: 20
Best features: ['attribute57', 'attribute21', 'attribute28', 'attribute18', 'attribute7', 'attribute53']




You can then repeat this for another 4 seeds and average the result to replicate the results in the paper.
We used seeds 42, 43, 44, 45, 46