Mercurial > repos > bgruening > sklearn_discriminant_classifier
view discriminant.xml @ 3:b6f35bef5b59 draft
planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 99231ce3ef238975a6614926f2fa6326b9454ee8
author | bgruening |
---|---|
date | Tue, 13 Mar 2018 06:31:58 -0400 |
parents | e0067d9baffc |
children | f46da2feb233 |
line wrap: on
line source
<tool id="sklearn_discriminant_classifier" name="Discriminant Analysis" version="@VERSION@"> <description></description> <macros> <import>main_macros.xml</import> <!--macro name="priors"--> </macros> <expand macro="python_requirements"/> <expand macro="macro_stdio"/> <version_command>echo "@VERSION@"</version_command> <command><![CDATA[ python "$discriminant_script" '$inputs' ]]> </command> <configfiles> <inputs name="inputs"/> <configfile name="discriminant_script"> <![CDATA[ import sys import json import numpy as np import sklearn.discriminant_analysis import pandas import pickle input_json_path = sys.argv[1] params = json.load(open(input_json_path, "r")) #if $selected_tasks.selected_task == "load": classifier_object = pickle.load(open("$infile_model", 'r')) data = pandas.read_csv("$selected_tasks.infile_data", sep='\t', header=0, index_col=None, parse_dates=True, encoding=None, tupleize_cols=False ) prediction = classifier_object.predict(data) prediction_df = pandas.DataFrame(prediction) res = pandas.concat([data, prediction_df], axis=1) res.to_csv(path_or_buf = "$outfile_predict", sep="\t", index=False) #else: data_train = pandas.read_csv("$selected_tasks.infile_train", sep='\t', header=0, index_col=None, parse_dates=True, encoding=None, tupleize_cols=False ) data = data_train.ix[:,0:len(data_train.columns)-1] labels = np.array(data_train[data_train.columns[len(data_train.columns)-1]]) options = params["selected_tasks"]["selected_algorithms"]["options"] selected_algorithm = params["selected_tasks"]["selected_algorithms"]["selected_algorithm"] my_class = getattr(sklearn.discriminant_analysis, selected_algorithm) classifier_object = my_class(**options) classifier_object.fit(data,labels) pickle.dump(classifier_object,open("$outfile_fit", 'w+'), pickle.HIGHEST_PROTOCOL) #end if ]]> </configfile> </configfiles> <inputs> <expand macro="train_loadConditional" model="zip"> <param name="selected_algorithm" type="select" label="Classifier type"> <option value="LinearDiscriminantAnalysis" selected="true">Linear Discriminant Classifier</option> <option value="QuadraticDiscriminantAnalysis">Quadratic Discriminant Classifier</option> </param> <when value="LinearDiscriminantAnalysis"> <section name="options" title="Advanced Options" expanded="False"> <param argument="solver" type="select" optional="true" label="Solver" help=""> <option value="svd" selected="true">Singular Value Decomposition</option> <option value="lsqr">Least Squares Solution</option> <option value="eigen">Eigenvalue Decomposition</option> </param> <!--param name="shrinkage"--> <!--expand macro="priors"/--> <param argument="n_components" type="integer" optional="true" value="" label="Number of components" help="Number of components for dimensionality reduction. ( always less than n_classes - 1 )"/> <expand macro="tol" default_value="0.0001" help_text="Rank estimation threshold used in SVD solver."/> <param argument="store_covariance" type="boolean" optional="true" truevalue="booltrue" falsevalue="boolflase" checked="false" label="Store covariance" help="Compute class covariance matrix."/> </section> </when> <when value="QuadraticDiscriminantAnalysis"> <section name="options" title="Advanced Options" expanded="False"> <!--expand macro="priors"/--> <param argument="reg_param" type="float" optional="true" value="0.0" label="Regularization coefficient" help="Covariance estimate regularizer."/> <expand macro="tol" default_value="0.00001" help_text="Rank estimation threshold used in SVD solver."/> <param argument="store_covariances" type="boolean" optional="true" truevalue="booltrue" falsevalue="boolflase" checked="false" label="Store covariances" help="Compute class covariance matrixes."/> </section> </when> </expand> </inputs> <expand macro="output"/> <tests> <test> <param name="infile_train" value="train.tabular" ftype="tabular"/> <param name="selected_task" value="train"/> <param name="selected_algorithm" value="LinearDiscriminantAnalysis"/> <param name="solver" value="svd" /> <param name="store_covariances" value="True"/> <output name="outfile_fit" file="lda_model01" compare="sim_size" delta="500"/> </test> <test> <param name="infile_train" value="train.tabular" ftype="tabular"/> <param name="selected_task" value="train"/> <param name="selected_algorithm" value="LinearDiscriminantAnalysis"/> <param name="solver" value="lsqr"/> <output name="outfile_fit" file="lda_model02" compare="sim_size" delta="500"/> </test> <test> <param name="infile_train" value="train.tabular" ftype="tabular"/> <param name="selected_task" value="train"/> <param name="selected_algorithm" value="QuadraticAnalysis"/> <output name="outfile_fit" file="qda_model01" compare="sim_size" delta="500"/> </test> <test> <param name="infile_model" value="lda_model01" ftype="zip"/> <param name="infile_data" value="test.tabular" ftype="tabular"/> <param name="selected_task" value="load"/> <output name="outfile_predict" file="lda_prediction_result01.tabular"/> </test> <test> <param name="infile_model" value="lda_model02" ftype="zip"/> <param name="infile_data" value="test.tabular" ftype="tabular"/> <param name="selected_task" value="load"/> <output name="outfile_predict" file="lda_prediction_result02.tabular"/> </test> <test> <param name="infile_model" value="qda_model01" ftype="zip"/> <param name="infile_data" value="test.tabular" ftype="tabular"/> <param name="selected_task" value="load"/> <output name="outfile_predict" file="qda_prediction_result01.tabular"/> </test> </tests> <help><![CDATA[ ***What it does*** Linear and Quadratic Discriminant Analysis are two classic classifiers with a linear and a quadratic decision surface respectively. These classifiers are fast and easy to interprete. **1 - Training input** When you choose to train a model, discriminant analysis tool expects a tabular file with numeric values, the order of the columns being as follows: :: "feature_1" "feature_2" "..." "feature_n" "class_label" **Example for training data** The following training dataset contains 3 feature columns and a column containing class labels: :: 4.01163365529 -6.10797684314 8.29829894763 1 10.0788438916 1.59539821454 10.0684278289 0 -5.17607775503 -0.878286135332 6.92941850665 2 4.00975406235 -7.11847496542 9.3802423585 1 4.61204065139 -5.71217537352 9.12509610964 1 **2 - Trainig output** Based on your choice, this tool fits a sklearn discriminant_analysis.LinearDiscriminantAnalysis or discriminant_analysis.QuadraticDiscriminantAnalysis on the traning data and outputs the trained model in the form of pickled object in a text file. **3 - Prediction input** When you choose to load a model and do prediction, the tool expects an already trained Discriminant Analysis estimator and a tabular dataset as input. The dataset is a tabular file with new samples which you want to classify. It just contains feature columns. **Example for prediction data** :: 8.26530668997 2.96705005011 8.88881190248 2.96366327113 -3.76295851562 11.7113372463 8.13319631944 -0.223645298585 10.5820605308 .. class:: warningmark The number of feature columns must be the same in training and prediction datasets! **3 - Prediction output** The tool predicts the class labels for new samples and adds them as the last column to the prediction dataset. The new dataset then is output as a tabular file. The prediction output format should look like the training dataset. Discriminant Analysis is based on sklearn.discriminant_analysis library from Scikit-learn. For more information please refer to `Scikit-learn site`_. .. _`Scikit-learn site`: http://scikit-learn.org/stable/modules/lda_qda.html ]]></help> <expand macro="sklearn_citation"/> </tool>