annotate featureselect/toolrfe.xml @ 4:5364cf43a8c1 draft

Uploaded
author deepakjadmin
date Sun, 02 Oct 2016 05:36:30 -0400
parents 91c141c5efa6
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
3
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
1 <tool id="featureSelectR" name="Feature Selection" >
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
2 <description>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
3 This tool used for extract best feature subsets cantaining input data for model building.
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
4 </description>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
5 <!--command interpreter="bash">step3run.sh $file1 $model $output1 2>/dev/null </command-->
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
6 <requirements>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
7 <requirement type="set_environment">FEATURE_SELECTION_R</requirement>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
8 <requirement type="set_environment">R_ROOT_DIR</requirement>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
9 <requirement type="package" version="3.2.0">R</requirement>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
10 <requirement type="package" version="1.0.0">caret-tools</requirement>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
11 </requirements>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
12 <command interpreter="Rscript">feature_selection.R $input $profile $finalset $function1 $resampling $repeat $number $corcutoff > /dev/null 2>&amp;1 </command>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
13
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
14 <inputs>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
15 <param name="input" type="data" label="Select input data file" help="input .RData file" />
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
16 <param name="function1" type="select" display="radio" label="Select appropriate function for algorithm" >
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
17 <option value="rfFuncs" selected="true">random forest based function </option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
18 <option value="lmFuncs">linear model based function</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
19 <option value="treebagFuncs">treebag(CART) based function</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
20 <option value="nbFuncs">neive bayes based function</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
21 </param>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
22
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
23 <param name="corcutoff" type="float" value= "0.8" min="0.0" max = "1.0" label="Select correlation cutoff" help="values bewteen 0-1. fileds above cufoff value removed from data " />
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
24 <param name="resampling" type="select" label="Select appropriate resampling method" >
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
25 <option value="repeatedcv" selected="true">repeatedcv </option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
26 <option value="boot">boot</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
27 <option value="cv">cv</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
28 <option value="boot632">boot632</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
29 </param>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
30 <param name="repeat" type="select" label="Set Number of times to repeat" help="default is 3 ">
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
31 <option value="3" selected="true">3</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
32 <option value="1">1</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
33 <option value="5">5</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
34 <option value="10">10</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
35 </param>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
36 <param name="number" type="select" label="Set Number of times Resample" help="default is 10">
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
37 <option value="10" selected="true">10</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
38 <option value="5">5</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
39 <option value="15">15</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
40 <option value="20">20</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
41 <option value="25">25</option>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
42 </param>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
43
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
44 </inputs>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
45 <outputs>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
46 <data type="data" format="data" name="profile" label="$function1-profile" />
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
47 <data type= "data" format="data" name="finalset" label="Selected_feature.RData "/>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
48 </outputs>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
49 <help>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
50 .. class:: infomark
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
51
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
52 **RFE based feature selection for classification and regression**
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
53
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
54 Input file must be RData file obtained by converting csv file in to RData.
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
55
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
56 output "Selected_feature.RData" file used for model building purpose.While profile
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
57
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
58 represents feature selection model.
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
59
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
60 Correlation cutoff value is desired for choosing independent variables For example
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
61
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
62 Cutoff value = 0.8 removes all descriptors sharing equal or highet correlation values.
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
63
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
64 User may choose varous resampling methods in combination with repeats and times of resample.
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
65
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
66
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
67
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
68 </help>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
69
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
70
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
71 <tests>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
72 <test>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
73 <param name="input" value="testinput.RData"/>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
74 <param name="function1" value="rfFuncs" />
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
75 <param name="corcutoff" value="0.6" />
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
76 <param name="resampling" value="repeatedcv" />
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
77 <param name="repeat" value="1" />
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
78 <param name="number" value="5" />
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
79
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
80
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
81 <output name="profile" file="rfprofile.RData" compare="sim_size" delta="2000000" />
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
82 <output name="finalset" file="selected_fet.RData" compare="sim_size" delta="2000000"/>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
83 </test>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
84 </tests>
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
85
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
86
91c141c5efa6 Uploaded
deepakjadmin
parents:
diff changeset
87 </tool>