
D:\DM\mRMR>mrmr_win32.exe

Usage: mrmr_osx -i <dataset> -t <threshold> [optional arguments]
         -i <dataset>    .CSV file containing M rows and N columns, row - sample, column - variable/attribute.
         -t <threshold> a float number of the discretization threshold; non-specifying this parameter means no discretizaton (i.e. data is already integer); 0 to make binarization.
         -n <number of features>   a natural number, default is 50.
         -m <selection method>    either "MID" or "MIQ" (Capital case), default is MID.
         -s <MAX number of samples>   a natural number, default is 1000. Note that if you don't have or don't need big memory, set this value small, as this program will use this value to pre-allocate memory in data file reading.
         -v <MAX number of variables/attibutes in data>   a natural number, default is 10000. Note that if you don't have or don't need big memory, set this value small, as this program will use this value to pre-allocate memory in data file reading.
         [-h] print this message.


 *** This program and the respective minimum Redundancy Maximum Relevance (mRMR)
     algorithm were developed by Hanchuan Peng <hanchuan.peng@gmail.com>for
     the paper
     "Feature selection based on mutual information: criteria of
      max-dependency, max-relevance, and min-redundancy,"
      Hanchuan Peng, Fuhui Long, and Chris Ding,
      IEEE Transactions on Pattern Analysis and Machine Intelligence,
      Vol. 27, No. 8, pp.1226-1238, 2005.

D:\DM\mRMR>mrmr_win32.exe -i test_nci9_s3.csv
You have specified parameters: no discretization, #fea=50 selection method=MID #maxVar=10000 #maxSample=1000

Target classification variable (#1 column in the input data) has name=class     entropy score=3.085

*** MaxRel features ***
Order    Fea     Name    Score
1        444     v444    0.900
2        7675    v7675   0.837
3        3534    v3534   0.762
4        9576    v9576   0.760
5        9577    v9577   0.748
6        5642    v5642   0.743
7        9588    v9588   0.742
8        7866    v7866   0.732
9        756     v756    0.732
10       1370    v1370   0.728
11       6291    v6291   0.723
12       1916    v1916   0.710
13       1700    v1700   0.709
14       7333    v7333   0.708
15       3484    v3484   0.703
16       7676    v7676   0.702
17       6934    v6934   0.694
18       6990    v6990   0.686
19       812     v812    0.683
20       4520    v4520   0.680
21       9400    v9400   0.680
22       8395    v8395   0.678
23       4938    v4938   0.677
24       3521    v3521   0.673
25       7472    v7472   0.671
26       6608    v6608   0.670
27       7332    v7332   0.667
28       5248    v5248   0.665
29       8363    v8363   0.663
30       8934    v8934   0.661
31       4073    v4073   0.659
32       635     v635    0.656
33       9394    v9394   0.656
34       1705    v1705   0.656
35       9226    v9226   0.655
36       8468    v8468   0.654
37       8195    v8195   0.650
38       5709    v5709   0.648
39       3485    v3485   0.645
40       9342    v9342   0.645
41       9320    v9320   0.645
42       6958    v6958   0.644
43       3641    v3641   0.643
44       744     v744    0.642
45       3676    v3676   0.640
46       262     v262    0.639
47       5131    v5131   0.638
48       3548    v3548   0.637
49       8088    v8088   0.631
50       8574    v8574   0.630

*** mRMR features ***
Order    Fea     Name    Score
1        444     v444    0.900
2        5642    v5642   0.735
3        756     v756    0.679
4        7675    v7675   0.676
5        9577    v9577   0.625
6        1700    v1700   0.623
7        3484    v3484   0.603
8        1370    v1370   0.591
9        6291    v6291   0.590
10       9400    v9400   0.573
11       3534    v3534   0.578
12       812     v812    0.564
13       9576    v9576   0.574
14       744     v744    0.564
15       3485    v3485   0.567
16       3521    v3521   0.547
17       6934    v6934   0.544
18       1916    v1916   0.542
19       4520    v4520   0.546
20       6608    v6608   0.540
21       8395    v8395   0.536
22       7866    v7866   0.530
23       5131    v5131   0.520
24       9588    v9588   0.521
25       6706    v6706   0.513
26       7333    v7333   0.517
27       512     v512    0.512
28       9342    v9342   0.508
29       6990    v6990   0.505
30       4938    v4938   0.507
31       9226    v9226   0.506
32       8935    v8935   0.502
33       456     v456    0.499
34       5248    v5248   0.500
35       262     v262    0.500
36       9320    v9320   0.498
37       4340    v4340   0.494
38       3676    v3676   0.492
39       8169    v8169   0.493
40       1466    v1466   0.492
41       7676    v7676   0.494
42       1705    v1705   0.488
43       3394    v3394   0.486
44       635     v635    0.486
45       5709    v5709   0.487
46       2160    v2160   0.486
47       7171    v7171   0.485
48       8698    v8698   0.485
49       2790    v2790   0.484
50       4638    v4638   0.485


 *** This program and the respective minimum Redundancy Maximum Relevance (mRMR)
     algorithm were developed by Hanchuan Peng <hanchuan.peng@gmail.com>for
     the paper
     "Feature selection based on mutual information: criteria of
      max-dependency, max-relevance, and min-redundancy,"
      Hanchuan Peng, Fuhui Long, and Chris Ding,
      IEEE Transactions on Pattern Analysis and Machine Intelligence,
      Vol. 27, No. 8, pp.1226-1238, 2005.
