Mercurial > repos > greg > gregs_test_repo
view CADDSuite-1.0.1/galaxyconfigs/tools/FeatureSelector.xml @ 26:2f19ad95a19d
Uploaded https://github.com/downloads/CADDSuite/main/CADDSuite_1.0.1_Linux_x86-64.tar.gz
author | greg |
---|---|
date | Tue, 15 Nov 2011 11:55:29 -0500 |
parents | 389056b8864a |
children |
line wrap: on
line source
<!--This is a configuration file for the integration of a CADDSuite tool into Galaxy (http://usegalaxy.org). This file was automatically generated using GalaxyConfigGenerator, so do not bother to make too many manual modifications.--> <tool id="featureselector" name="FeatureSelector" version="1.0.1"> <description>run feature-selection on a QSAR model</description> <command interpreter="bash"><![CDATA[../../FeatureSelector #if str( $i ) != '' and str( $i ) != 'None' : -i "$i" #end if #if str( $dat ) != '' and str( $dat ) != 'None' : -dat "$dat" #end if #if str( $o ) != '' and str( $o ) != 'None' : -o "$o" #end if #if str( $type ) != '' and str( $type ) != 'None' : -type "$type" #end if | tail -n 5 ]]></command> <inputs> <param name="i" optional="false" label="input mod-file" type="data" format="mod"/> <param name="dat" optional="false" label="data-file" type="data" format="dat"/> <param name="type" optional="false" label="feature-selection type" type="select"> <option value="remove_correlated">remove_correlated</option> <option value="forward_selection">forward_selection</option> <option value="backward_selection">backward_selection</option> <option value="stepwise_selection">stepwise_selection</option> <option value="twinscan">twinscan</option> <option value="removeLowResponseCorrelation">removeLowResponseCorrelation</option> </param> </inputs> <outputs> <data name="o" format="mod"/> </outputs> <help>FeatureSelector runs a feature-selection for a given QSAR model. The type of feature-selection to be done is specified by '-type'. Input of this tool is a data file as generated by InputReader containing the training data for feature-selection and a QSAR model file as generated by ModelCreator (or this tool itself). Note that you can apply several feature-selection methods in succession by using the output of one call of this tool as input for the next call. Model- and kernel-parameters (if any) will be automatically optimized by cross-validation after applying the desired feature-selection. Output of this tool is a model-file that can be used by other QuEasy tools (e.g. Validator).</help> </tool>