Mercurial > repos > bgruening > sklearn_data_preprocess

comparison pre_process.xml @ 35:0e5fcf7ddc75 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit e2a5eade6d0e5ddf3a47630381a0ad90d80e8a04"
author bgruening
date Tue, 13 Apr 2021 18:53:33 +0000
parents eb79bde99328
children a16f33c6ca64
comparison
equal deleted inserted replaced
34:eb9da067ab26 35:0e5fcf7ddc75
1 <tool id="sklearn_data_preprocess" name="Preprocess" version="@VERSION@"> 1 <tool id="sklearn_data_preprocess" name="Preprocess" version="@VERSION@" profile="20.05">
2 <description>raw feature vectors into standardized datasets</description> 2 <description>raw feature vectors into standardized datasets</description>
3 <macros> 3 <macros>
4 <import>main_macros.xml</import> 4 <import>main_macros.xml</import>
5 </macros> 5 </macros>
6 <expand macro="python_requirements"/> 6 <expand macro="python_requirements" />
7 <expand macro="macro_stdio"/> 7 <expand macro="macro_stdio" />
8 <version_command>echo "@VERSION@"</version_command> 8 <version_command>echo "@VERSION@"</version_command>
9 <command> 9 <command>
10 <![CDATA[ 10 <![CDATA[
11 python "$pre_processor_script" '$inputs' 11 python "$pre_processor_script" '$inputs'
12 ]]> 12 ]]>
95 </param> 95 </param>
96 <when value="tabular"> 96 <when value="tabular">
97 <param name="infile" type="data" format="tabular" label="Select a tabular file you want to train your preprocessor on its data:" /> 97 <param name="infile" type="data" format="tabular" label="Select a tabular file you want to train your preprocessor on its data:" />
98 <param name="header1" type="boolean" optional="true" truevalue="booltrue" falsevalue="boolfalse" checked="false" label="Does the dataset contain header:" /> 98 <param name="header1" type="boolean" optional="true" truevalue="booltrue" falsevalue="boolfalse" checked="false" label="Does the dataset contain header:" />
99 <conditional name="column_selector_options_1"> 99 <conditional name="column_selector_options_1">
100 <expand macro="samples_column_selector_options" multiple="true" column_option="selected_column_selector_option" col_name="col1" infile="infile"/> 100 <expand macro="samples_column_selector_options" multiple="true" column_option="selected_column_selector_option" col_name="col1" infile="infile" />
101 </conditional> 101 </conditional>
102 <conditional name="pre_processors"> 102 <conditional name="pre_processors">
103 <expand macro="sparse_preprocessors_ext" /> 103 <expand macro="sparse_preprocessors_ext" />
104 <expand macro="sparse_preprocessor_options_ext" /> 104 <expand macro="sparse_preprocessor_options_ext" />
105 </conditional> 105 </conditional>
106 </when> 106 </when>
107 <when value="sparse"> 107 <when value="sparse">
108 <param name="infile" type="data" format="txt" label="Select a sparse representation you want to train your preprocessor on its data:"/> 108 <param name="infile" type="data" format="txt" label="Select a sparse representation you want to train your preprocessor on its data:" />
109 <conditional name="pre_processors"> 109 <conditional name="pre_processors">
110 <expand macro="sparse_preprocessors"/> 110 <expand macro="sparse_preprocessors" />
111 <expand macro="sparse_preprocessor_options"/> 111 <expand macro="sparse_preprocessor_options" />
112 </conditional> 112 </conditional>
113 </when> 113 </when>
114 </conditional> 114 </conditional>
115 <param name="save" type="boolean" truevalue="booltrue" falsevalue="boolflase" checked="false" 115 <param name="save" type="boolean" truevalue="booltrue" falsevalue="boolflase" checked="false" label="Save the preprocessor" help="Saves the preprocessor after fitting to the data. The preprocessor can then be passed to other tools and used in later operations." />
116 label="Save the preprocessor"
117 help="Saves the preprocessor after fitting to the data. The preprocessor can then be passed to other tools and used in later operations."/>
118 </inputs> 116 </inputs>
119 <outputs> 117 <outputs>
120 <data format="tabular" name="outfile_transform" from_work_dir="./output"/> 118 <data format="tabular" name="outfile_transform" from_work_dir="./output" />
121 <data format="zip" name="outfile_fit"> 119 <data format="zip" name="outfile_fit">
122 <filter>save</filter> 120 <filter>save</filter>
123 </data> 121 </data>
124 </outputs> 122 </outputs>
125 <tests> 123 <tests>
126 <test> 124 <test>
127 <param name="infile" value="train.tabular" ftype="tabular"/> 125 <param name="infile" value="train.tabular" ftype="tabular" />
128 <param name="selected_column_selector_option" value="all_columns"/> 126 <param name="selected_column_selector_option" value="all_columns" />
129 <param name="selected_input_type" value="tabular"/> 127 <param name="selected_input_type" value="tabular" />
130 <param name="selected_pre_processor" value="KernelCenterer"/> 128 <param name="selected_pre_processor" value="KernelCenterer" />
131 <param name="save" value="true"/> 129 <param name="save" value="true" />
132 <output name="outfile_transform" file="prp_result01" ftype="tabular"/> 130 <output name="outfile_transform" file="prp_result01" ftype="tabular" />
133 <output name="outfile_fit" file="prp_model01" ftype="zip" compare="sim_size" delta="5"/> 131 <output name="outfile_fit" file="prp_model01" ftype="zip" compare="sim_size" delta="5" />
134 </test> 132 </test>
135 <test> 133 <test>
136 <param name="infile" value="train.tabular" ftype="tabular"/> 134 <param name="infile" value="train.tabular" ftype="tabular" />
137 <param name="selected_column_selector_option" value="all_columns"/> 135 <param name="selected_column_selector_option" value="all_columns" />
138 <param name="selected_input_type" value="tabular"/> 136 <param name="selected_input_type" value="tabular" />
139 <param name="selected_pre_processor" value="MinMaxScaler"/> 137 <param name="selected_pre_processor" value="MinMaxScaler" />
140 <param name="save" value="true"/> 138 <param name="save" value="true" />
141 <output name="outfile_transform" file="prp_result02" ftype="tabular"/> 139 <output name="outfile_transform" file="prp_result02" ftype="tabular" />
142 <output name="outfile_fit" file="prp_model02" ftype="zip" compare="sim_size" delta="5"/> 140 <output name="outfile_fit" file="prp_model02" ftype="zip" compare="sim_size" delta="5" />
143 </test> 141 </test>
144 <test> 142 <test>
145 <param name="infile" value="train.tabular" ftype="tabular"/> 143 <param name="infile" value="train.tabular" ftype="tabular" />
146 <param name="selected_column_selector_option" value="all_columns"/> 144 <param name="selected_column_selector_option" value="all_columns" />
147 <param name="selected_input_type" value="tabular"/> 145 <param name="selected_input_type" value="tabular" />
148 <param name="selected_pre_processor" value="PolynomialFeatures"/> 146 <param name="selected_pre_processor" value="PolynomialFeatures" />
149 <param name="save" value="true"/> 147 <param name="save" value="true" />
150 <output name="outfile_transform" file="prp_result03" ftype="tabular"/> 148 <output name="outfile_transform" file="prp_result03" ftype="tabular" />
151 <output name="outfile_fit" file="prp_model03" ftype="zip" compare="sim_size" delta="5"/> 149 <output name="outfile_fit" file="prp_model03" ftype="zip" compare="sim_size" delta="5" />
152 </test> 150 </test>
153 <test> 151 <test>
154 <param name="infile" value="train.tabular" ftype="tabular"/> 152 <param name="infile" value="train.tabular" ftype="tabular" />
155 <param name="selected_column_selector_option" value="all_columns"/> 153 <param name="selected_column_selector_option" value="all_columns" />
156 <param name="selected_input_type" value="tabular"/> 154 <param name="selected_input_type" value="tabular" />
157 <param name="selected_pre_processor" value="RobustScaler"/> 155 <param name="selected_pre_processor" value="RobustScaler" />
158 <param name="save" value="true"/> 156 <param name="save" value="true" />
159 <output name="outfile_transform" file="prp_result04" ftype="tabular"/> 157 <output name="outfile_transform" file="prp_result04" ftype="tabular" />
160 <output name="outfile_fit" file="prp_model04" ftype="zip" compare="sim_size" delta="5"/> 158 <output name="outfile_fit" file="prp_model04" ftype="zip" compare="sim_size" delta="5" />
161 </test> 159 </test>
162 <test> 160 <test>
163 <param name="infile" value="csr_sparse2.mtx" ftype="txt"/> 161 <param name="infile" value="csr_sparse2.mtx" ftype="txt" />
164 <param name="selected_input_type" value="sparse"/> 162 <param name="selected_input_type" value="sparse" />
165 <param name="selected_pre_processor" value="Binarizer"/> 163 <param name="selected_pre_processor" value="Binarizer" />
166 <param name="save" value="true"/> 164 <param name="save" value="true" />
167 <output name="outfile_transform" file="prp_result05" ftype="tabular"/> 165 <output name="outfile_transform" file="prp_result05" ftype="tabular" />
168 <output name="outfile_fit" file="prp_model05" ftype="zip" compare="sim_size" delta="5"/> 166 <output name="outfile_fit" file="prp_model05" ftype="zip" compare="sim_size" delta="5" />
169 </test> 167 </test>
170 <test> 168 <test>
171 <param name="infile" value="train.tabular" ftype="tabular"/> 169 <param name="infile" value="train.tabular" ftype="tabular" />
172 <param name="selected_input_type" value="tabular"/> 170 <param name="selected_input_type" value="tabular" />
173 <param name="selected_column_selector_option" value="all_columns"/> 171 <param name="selected_column_selector_option" value="all_columns" />
174 <param name="selected_pre_processor" value="StandardScaler"/> 172 <param name="selected_pre_processor" value="StandardScaler" />
175 <param name="save" value="true"/> 173 <param name="save" value="true" />
176 <output name="outfile_transform" file="prp_result07" ftype="tabular"/> 174 <output name="outfile_transform" file="prp_result07" ftype="tabular" />
177 <output name="outfile_fit" file="prp_model07" ftype="zip" compare="sim_size" delta="5"/> 175 <output name="outfile_fit" file="prp_model07" ftype="zip" compare="sim_size" delta="5" />
178 </test> 176 </test>
179 <test> 177 <test>
180 <param name="infile" value="csr_sparse2.mtx" ftype="txt"/> 178 <param name="infile" value="csr_sparse2.mtx" ftype="txt" />
181 <param name="selected_input_type" value="sparse"/> 179 <param name="selected_input_type" value="sparse" />
182 <param name="selected_pre_processor" value="MaxAbsScaler"/> 180 <param name="selected_pre_processor" value="MaxAbsScaler" />
183 <param name="save" value="true"/> 181 <param name="save" value="true" />
184 <output name="outfile_transform" file="prp_result08" ftype="tabular"/> 182 <output name="outfile_transform" file="prp_result08" ftype="tabular" />
185 <output name="outfile_fit" file="prp_model08" ftype="zip" compare="sim_size" delta="5"/> 183 <output name="outfile_fit" file="prp_model08" ftype="zip" compare="sim_size" delta="5" />
186 </test> 184 </test>
187 <test> 185 <test>
188 <param name="infile" value="csr_sparse2.mtx" ftype="txt"/> 186 <param name="infile" value="csr_sparse2.mtx" ftype="txt" />
189 <param name="selected_input_type" value="sparse"/> 187 <param name="selected_input_type" value="sparse" />
190 <param name="selected_pre_processor" value="Normalizer"/> 188 <param name="selected_pre_processor" value="Normalizer" />
191 <param name="save" value="true"/> 189 <param name="save" value="true" />
192 <output name="outfile_transform" file="prp_result09" ftype="tabular"/> 190 <output name="outfile_transform" file="prp_result09" ftype="tabular" />
193 <output name="outfile_fit" file="prp_model09" ftype="zip" compare="sim_size" delta="5"/> 191 <output name="outfile_fit" file="prp_model09" ftype="zip" compare="sim_size" delta="5" />
194 </test> 192 </test>
195 <test> 193 <test>
196 <param name="infile" value="regression_X.tabular" ftype="tabular"/> 194 <param name="infile" value="regression_X.tabular" ftype="tabular" />
197 <param name="header1" value="true"/> 195 <param name="header1" value="true" />
198 <param name="selected_column_selector_option" value="all_columns"/> 196 <param name="selected_column_selector_option" value="all_columns" />
199 <param name="selected_input_type" value="tabular"/> 197 <param name="selected_input_type" value="tabular" />
200 <param name="selected_pre_processor" value="MinMaxScaler"/> 198 <param name="selected_pre_processor" value="MinMaxScaler" />
201 <param name="feature_range" value="(-1, 1)"/> 199 <param name="feature_range" value="(-1, 1)" />
202 <param name="save" value="false"/> 200 <param name="save" value="false" />
203 <output name="outfile_transform" file="prp_result10" ftype="tabular"/> 201 <output name="outfile_transform" file="prp_result10" ftype="tabular" />
204 </test> 202 </test>
205 </tests> 203 </tests>
206 <help> 204 <help>
207 <![CDATA[ 205 <![CDATA[
208 **What it does** 206 **What it does**
212 For information about preprocessing classes and parameter settings please refer to `Scikit-learn preprocessing`_. 210 For information about preprocessing classes and parameter settings please refer to `Scikit-learn preprocessing`_.
213 211
214 .. _`Scikit-learn preprocessing`: http://scikit-learn.org/stable/modules/preprocessing.html 212 .. _`Scikit-learn preprocessing`: http://scikit-learn.org/stable/modules/preprocessing.html
215 ]]> 213 ]]>
216 </help> 214 </help>
217 <expand macro="sklearn_citation"/> 215 <expand macro="sklearn_citation" />
218 </tool> 216 </tool>