List of usage examples for org.apache.mahout.classifier ResultAnalyzer addInstance
public boolean addInstance(String correctLabel, ClassifierResult classifiedResult)
From source file:com.memonews.mahout.sentiment.SentimentModelTester.java
License:Apache License
public void run(final PrintWriter output) throws IOException { final File base = new File(inputFile); // contains the best model final OnlineLogisticRegression classifier = ModelSerializer.readBinary(new FileInputStream(modelFile), OnlineLogisticRegression.class); final Dictionary newsGroups = new Dictionary(); final Multiset<String> overallCounts = HashMultiset.create(); final List<File> files = Lists.newArrayList(); for (final File newsgroup : base.listFiles()) { if (newsgroup.isDirectory()) { newsGroups.intern(newsgroup.getName()); files.addAll(Arrays.asList(newsgroup.listFiles())); }/*from ww w .ja v a 2s.c o m*/ } System.out.printf("%d test files\n", files.size()); final ResultAnalyzer ra = new ResultAnalyzer(newsGroups.values(), "DEFAULT"); for (final File file : files) { final String ng = file.getParentFile().getName(); final int actual = newsGroups.intern(ng); final SentimentModelHelper helper = new SentimentModelHelper(); final Vector input = helper.encodeFeatureVector(file, overallCounts);// no // leak // type // ensures // this // is // a // normal // vector final Vector result = classifier.classifyFull(input); final int cat = result.maxValueIndex(); final double score = result.maxValue(); final double ll = classifier.logLikelihood(actual, input); final ClassifierResult cr = new ClassifierResult(newsGroups.values().get(cat), score, ll); ra.addInstance(newsGroups.values().get(actual), cr); } output.printf("%s\n\n", ra.toString()); }
From source file:com.missionsky.scp.dataanalysis.mahout.TestNaiveBayesDriver.java
License:Apache License
private static void analyzeResults(Map<Integer, String> labelMap, SequenceFileDirIterable<Text, VectorWritable> dirIterable, ResultAnalyzer analyzer) { for (Pair<Text, VectorWritable> pair : dirIterable) { int bestIdx = Integer.MIN_VALUE; double bestScore = Long.MIN_VALUE; for (Vector.Element element : pair.getSecond().get().all()) { if (element.get() > bestScore) { bestScore = element.get(); bestIdx = element.index(); }//from w ww . j a v a2 s . com } if (bestIdx != Integer.MIN_VALUE) { ClassifierResult classifierResult = new ClassifierResult(labelMap.get(bestIdx), bestScore); analyzer.addInstance(pair.getFirst().toString(), classifierResult); } } }
From source file:com.tamingtext.classifier.maxent.TestMaxent.java
License:Apache License
private static void runTest(File[] inputFiles, DocumentCategorizer categorizer, Tokenizer tokenizer, ResultAnalyzer resultAnalyzer) throws FileNotFoundException, IOException { String line;//ww w. j ava 2 s. c o m //<start id="maxent.examples.test.execute"/> for (File ff : inputFiles) { BufferedReader in = new BufferedReader(new FileReader(ff)); while ((line = in.readLine()) != null) { String[] parts = line.split("\t"); if (parts.length != 2) continue; String docText = parts[1]; //<co id="tmt.preprocess"/> String[] tokens = tokenizer.tokenize(docText); double[] probs = categorizer.categorize(tokens); //<co id="tmt.categorize"/> String label = categorizer.getBestCategory(probs); int bestIndex = categorizer.getIndex(label); double score = probs[bestIndex]; ClassifierResult result //<co id="tmt.collect"/> = new ClassifierResult(label, score); resultAnalyzer.addInstance(parts[0], result); } in.close(); } System.err.println(resultAnalyzer.toString()); //<co id="tmt.summarize"/> /*<calloutlist> * <callout arearefs="tmt.preprocess">Preprocess text</callout> * <callout arearefs="tmt.categorize">Categorize</callout> * <callout arearefs="tmt.collect">Analyze Results</callout> * <callout arearefs="tmt.summarize">Present Results</callout> * </calloutlist>*/ //<end id="maxent.examples.test.execute"/> }
From source file:com.tamingtext.classifier.mlt.TestMoreLikeThis.java
License:Apache License
public static void main(String[] args) throws Exception { DefaultOptionBuilder obuilder = new DefaultOptionBuilder(); ArgumentBuilder abuilder = new ArgumentBuilder(); GroupBuilder gbuilder = new GroupBuilder(); Option helpOpt = DefaultOptionCreator.helpOption(); Option inputDirOpt = obuilder.withLongName("input").withRequired(true) .withArgument(abuilder.withName("input").withMinimum(1).withMaximum(1).create()) .withDescription("The input directory").withShortName("i").create(); Option modelOpt = obuilder.withLongName("model").withRequired(true) .withArgument(abuilder.withName("index").withMinimum(1).withMaximum(1).create()) .withDescription("The directory containing the index model").withShortName("m").create(); Option categoryFieldOpt = obuilder.withLongName("categoryField").withRequired(true) .withArgument(abuilder.withName("index").withMinimum(1).withMaximum(1).create()) .withDescription("Name of the field containing category information").withShortName("catf") .create();//from w ww .j a v a 2s . co m Option contentFieldOpt = obuilder.withLongName("contentField").withRequired(true) .withArgument(abuilder.withName("index").withMinimum(1).withMaximum(1).create()) .withDescription("Name of the field containing content information").withShortName("contf") .create(); Option maxResultsOpt = obuilder.withLongName("maxResults").withRequired(false) .withArgument(abuilder.withName("gramSize").withMinimum(1).withMaximum(1).create()) .withDescription("Number of results to retrive, default: 10 ").withShortName("r").create(); Option gramSizeOpt = obuilder.withLongName("gramSize").withRequired(false) .withArgument(abuilder.withName("gramSize").withMinimum(1).withMaximum(1).create()) .withDescription("Size of the n-gram. Default Value: 1 ").withShortName("ng").create(); Option typeOpt = obuilder.withLongName("classifierType").withRequired(false) .withArgument(abuilder.withName("classifierType").withMinimum(1).withMaximum(1).create()) .withDescription("Type of classifier: knn|tfidf. Default: bayes").withShortName("type").create(); Group group = gbuilder.withName("Options").withOption(gramSizeOpt).withOption(helpOpt) .withOption(inputDirOpt).withOption(modelOpt).withOption(typeOpt).withOption(contentFieldOpt) .withOption(categoryFieldOpt).withOption(maxResultsOpt).create(); try { Parser parser = new Parser(); parser.setGroup(group); parser.setHelpOption(helpOpt); CommandLine cmdLine = parser.parse(args); if (cmdLine.hasOption(helpOpt)) { CommandLineUtil.printHelp(group); return; } String classifierType = (String) cmdLine.getValue(typeOpt); int gramSize = 1; if (cmdLine.hasOption(gramSizeOpt)) { gramSize = Integer.parseInt((String) cmdLine.getValue(gramSizeOpt)); } int maxResults = 10; if (cmdLine.hasOption(maxResultsOpt)) { maxResults = Integer.parseInt((String) cmdLine.getValue(maxResultsOpt)); } String inputPath = (String) cmdLine.getValue(inputDirOpt); String modelPath = (String) cmdLine.getValue(modelOpt); String categoryField = (String) cmdLine.getValue(categoryFieldOpt); String contentField = (String) cmdLine.getValue(contentFieldOpt); MatchMode mode; if ("knn".equalsIgnoreCase(classifierType)) { mode = MatchMode.KNN; } else if ("tfidf".equalsIgnoreCase(classifierType)) { mode = MatchMode.TFIDF; } else { throw new IllegalArgumentException("Unkown classifierType: " + classifierType); } Directory directory = FSDirectory.open(new File(modelPath)); IndexReader indexReader = IndexReader.open(directory); Analyzer analyzer //<co id="mlt.analyzersetup"/> = new EnglishAnalyzer(Version.LUCENE_36); MoreLikeThisCategorizer categorizer = new MoreLikeThisCategorizer(indexReader, categoryField); categorizer.setAnalyzer(analyzer); categorizer.setMatchMode(mode); categorizer.setFieldNames(new String[] { contentField }); categorizer.setMaxResults(maxResults); categorizer.setNgramSize(gramSize); File f = new File(inputPath); if (!f.isDirectory()) { throw new IllegalArgumentException(f + " is not a directory or does not exit"); } File[] inputFiles = FileUtil.buildFileList(f); String line = null; //<start id="lucene.examples.mlt.test"/> final ClassifierResult UNKNOWN = new ClassifierResult("unknown", 1.0); ResultAnalyzer resultAnalyzer = //<co id="co.mlt.ra"/> new ResultAnalyzer(categorizer.getCategories(), UNKNOWN.getLabel()); for (File ff : inputFiles) { //<co id="co.mlt.read"/> BufferedReader in = new BufferedReader(new InputStreamReader(new FileInputStream(ff), "UTF-8")); while ((line = in.readLine()) != null) { String[] parts = line.split("\t"); if (parts.length != 2) { continue; } CategoryHits[] hits //<co id="co.mlt.cat"/> = categorizer.categorize(new StringReader(parts[1])); ClassifierResult result = hits.length > 0 ? hits[0] : UNKNOWN; resultAnalyzer.addInstance(parts[0], result); //<co id="co.mlt.an"/> } in.close(); } System.out.println(resultAnalyzer.toString());//<co id="co.mlt.print"/> /* <calloutlist> <callout arearefs="co.mlt.ra">Create <classname>ResultAnalyzer</classname></callout> <callout arearefs="co.mlt.read">Read Test data</callout> <callout arearefs="co.mlt.cat">Categorize</callout> <callout arearefs="co.mlt.an">Collect Results</callout> <callout arearefs="co.mlt.print">Display Results</callout> </calloutlist> */ //<end id="lucene.examples.mlt.test"/> } catch (OptionException e) { log.error("Error while parsing options", e); } }
From source file:com.wsc.myexample.decisionForest.MyTestForest.java
License:Apache License
private void testFile(String inPath, String outPath, DataConverter converter, MyDecisionForest forest, Dataset dataset, /*List<double[]> results,*/ Random rng, ResultAnalyzer analyzer) throws IOException { // create the predictions file DataOutputStream ofile = null; if (outPath != null) { ofile = new DataOutputStream(new FileOutputStream(outPath)); }/*ww w .ja va2s .co m*/ DataInputStream input = new DataInputStream(new FileInputStream(inPath)); try { Scanner scanner = new Scanner(input); while (scanner.hasNextLine()) { String line = scanner.nextLine(); if (line.isEmpty()) { continue; // skip empty lines } Instance instance = converter.convert(line); if (instance == null) continue; double prediction = forest.classify(dataset, rng, instance); if (ofile != null) { ofile.writeChars(Double.toString(prediction)); // write the prediction ofile.writeChar('\n'); } // results.add(new double[] {dataset.getLabel(instance), prediction}); analyzer.addInstance(dataset.getLabelString(dataset.getLabel(instance)), new ClassifierResult(dataset.getLabelString(prediction), 1.0)); } scanner.close(); } finally { Closeables.closeQuietly(input); ofile.close(); } }
From source file:guipart.view.GUIOverviewController.java
@FXML void handleClassifyRF(ActionEvent event) throws IOException { String outputFile = "data/out"; Path dataPath = new Path(textFieldCSVRF.getText()); // test data path Path datasetPath = new Path(textFieldDatasetRF.getText()); //info file about data set Path modelPath = new Path(textFieldModelRF.getText()); // path where the forest is stored Path outputPath = new Path(outputFile); // path to predictions file, if null do not output the predictions Configuration conf = new Configuration(); FileSystem fs = FileSystem.get(conf); FileSystem outFS = FileSystem.get(conf); System.out.println("Loading the forest"); DecisionForest forest = DecisionForest.load(conf, modelPath); if (forest == null) System.err.println("No decision forest found!"); // load the dataset Dataset dataset = Dataset.load(conf, datasetPath); DataConverter converter = new DataConverter(dataset); System.out.println("Sequential classification"); long time = System.currentTimeMillis(); Random rng = RandomUtils.getRandom(); List<double[]> resList = Lists.newArrayList(); if (fs.getFileStatus(dataPath).isDir()) { //the input is a directory of files Utils.rfTestDirectory(outputPath, converter, forest, dataset, resList, rng, fs, dataPath, outFS, guiPart);//from w w w.j a va 2 s. co m } else { // the input is one single file Utils.rfTestFile(dataPath, outputPath, converter, forest, dataset, resList, rng, outFS, fs, guiPart); } time = System.currentTimeMillis() - time; //log.info("Classification Time: {}", DFUtils.elapsedTime(time)); System.out.println("Classification time: " + DFUtils.elapsedTime(time)); if (dataset.isNumerical(dataset.getLabelId())) { RegressionResultAnalyzer regressionAnalyzer = new RegressionResultAnalyzer(); double[][] results = new double[resList.size()][2]; regressionAnalyzer.setInstances(resList.toArray(results)); //log.info("{}", regressionAnalyzer); System.out.println(regressionAnalyzer.toString()); } else { ResultAnalyzer analyzer = new ResultAnalyzer(Arrays.asList(dataset.labels()), "unknown"); for (double[] r : resList) { analyzer.addInstance(dataset.getLabelString(r[0]), new ClassifierResult(dataset.getLabelString(r[1]), 1.0)); } //log.info("{}", analyzer); System.out.println(analyzer.toString()); textAnalyze.setText(analyzer.toString()); } }
From source file:imageClassify.TestForest.java
License:Apache License
private void mapreduce() throws ClassNotFoundException, IOException, InterruptedException { if (outputPath == null) { throw new IllegalArgumentException( "You must specify the ouputPath when using the mapreduce implementation"); }/* w ww . jav a2s . c om*/ Classifier classifier = new Classifier(modelPath, dataPath, datasetPath, outputPath, getConf()); classifier.run(); if (analyze) { double[][] results = classifier.getResults(); if (results != null) { Dataset dataset = Dataset.load(getConf(), datasetPath); if (dataset.isNumerical(dataset.getLabelId())) { RegressionResultAnalyzer regressionAnalyzer = new RegressionResultAnalyzer(); regressionAnalyzer.setInstances(results); log.info("{}", regressionAnalyzer); } else { ResultAnalyzer analyzer = new ResultAnalyzer(Arrays.asList(dataset.labels()), "unknown"); for (double[] res : results) { analyzer.addInstance(dataset.getLabelString(res[0]), new ClassifierResult(dataset.getLabelString(res[1]), 1.0)); } log.info("{}", analyzer); } } } }
From source file:imageClassify.TestForest.java
License:Apache License
private void sequential() throws IOException { log.info("Loading the forest..."); DecisionForest forest = DecisionForest.load(getConf(), modelPath); if (forest == null) { log.error("No Decision Forest found!"); return;//from w w w .ja va2 s . c o m } // load the dataset Dataset dataset = Dataset.load(getConf(), datasetPath); DataConverter converter = new DataConverter(dataset); log.info("Sequential classification..."); long time = System.currentTimeMillis(); Random rng = RandomUtils.getRandom(); List<double[]> resList = Lists.newArrayList(); if (dataFS.getFileStatus(dataPath).isDir()) { //the input is a directory of files testDirectory(outputPath, converter, forest, dataset, resList, rng); } else { // the input is one single file testFile(dataPath, outputPath, converter, forest, dataset, resList, rng); } time = System.currentTimeMillis() - time; log.info("Classification Time: {}", DFUtils.elapsedTime(time)); if (analyze) { if (dataset.isNumerical(dataset.getLabelId())) { RegressionResultAnalyzer regressionAnalyzer = new RegressionResultAnalyzer(); double[][] results = new double[resList.size()][2]; regressionAnalyzer.setInstances(resList.toArray(results)); log.info("{}", regressionAnalyzer); } else { ResultAnalyzer analyzer = new ResultAnalyzer(Arrays.asList(dataset.labels()), "unknown"); for (double[] r : resList) { analyzer.addInstance(dataset.getLabelString(r[0]), new ClassifierResult(dataset.getLabelString(r[1]), 1.0)); } log.info("{}", analyzer); } } }
From source file:javaapplication3.runRandomForest.java
public static void main(String[] args) throws InterruptedException, IOException, ClassNotFoundException { String outputFile = "data/lule24"; String inputFile = "data/DataFraud1MTest.csv"; String modelFile = "data/forest.seq"; String infoFile = "data/DataFraud1M.info"; Path dataPath = new Path(inputFile); // test data path Path datasetPath = new Path(infoFile); Path modelPath = new Path(modelFile); // path where the forest is stored Path outputPath = new Path(outputFile); // path to predictions file, if null do not output the predictions Configuration conf = new Configuration(); FileSystem fs = FileSystem.get(conf); /*/*from ww w . java 2s . c o m*/ p = Runtime.getRuntime().exec("bash /home/ivan/hadoop-1.2.1/bin/start-all.sh"); p.waitFor();*/ if (outputPath == null) { throw new IllegalArgumentException( "You must specify the ouputPath when using the mapreduce implementation"); } Classifier classifier = new Classifier(modelPath, dataPath, datasetPath, outputPath, conf); classifier.run(); double[][] results = classifier.getResults(); if (results != null) { Dataset dataset = Dataset.load(conf, datasetPath); Data data = DataLoader.loadData(dataset, fs, dataPath); Instance inst; for (int i = 0; i < data.size(); i++) { inst = data.get(i); //System.out.println("Prediction:"+inst.get(7)+" Real value:"+results[i][1]); System.out.println(inst.get(0) + " " + inst.get(1) + " " + inst.get(2) + " " + inst.get(3) + " " + inst.get(4) + " " + inst.get(5) + " " + inst.get(6) + " " + inst.get(7) + " "); } ResultAnalyzer analyzer = new ResultAnalyzer(Arrays.asList(dataset.labels()), "unknown"); for (double[] res : results) { analyzer.addInstance(dataset.getLabelString(res[0]), new ClassifierResult(dataset.getLabelString(res[1]), 1.0)); System.out.println("Prvi shit:" + res[0] + " Drugi Shit" + res[1]); } System.out.println(analyzer.toString()); } }
From source file:javaapplication3.RunRandomForestSeq.java
public static void main(String[] args) throws IOException { String outputFile = "data/out"; String inputFile = "data/DataFraud1MTest.csv"; String modelFile = "data/forest.seq"; String infoFile = "data/DataFraud1M.info"; Path dataPath = new Path(inputFile); // test data path Path datasetPath = new Path(infoFile); Path modelPath = new Path(modelFile); // path where the forest is stored Path outputPath = new Path(outputFile); // path to predictions file, if null do not output the predictions Configuration conf = new Configuration(); FileSystem fs = FileSystem.get(conf); FileSystem outFS = FileSystem.get(conf); //log.info("Loading the forest..."); System.out.println("Loading the forest"); DecisionForest forest = DecisionForest.load(conf, modelPath); if (forest == null) System.err.println("No decision forest found!"); //log.error("No Decision Forest found!"); // load the dataset Dataset dataset = Dataset.load(conf, datasetPath); DataConverter converter = new DataConverter(dataset); //log.info("Sequential classification..."); System.out.println("Sequential classification"); long time = System.currentTimeMillis(); Random rng = RandomUtils.getRandom(); List<double[]> resList = Lists.newArrayList(); if (fs.getFileStatus(dataPath).isDir()) { //the input is a directory of files testDirectory(outputPath, converter, forest, dataset, resList, rng, fs, dataPath, outFS); } else {/*from w w w. j a v a 2 s . co m*/ // the input is one single file testFile(dataPath, outputPath, converter, forest, dataset, resList, rng, outFS, fs); } time = System.currentTimeMillis() - time; //log.info("Classification Time: {}", DFUtils.elapsedTime(time)); System.out.println("Classification time: " + DFUtils.elapsedTime(time)); if (dataset.isNumerical(dataset.getLabelId())) { RegressionResultAnalyzer regressionAnalyzer = new RegressionResultAnalyzer(); double[][] results = new double[resList.size()][2]; regressionAnalyzer.setInstances(resList.toArray(results)); //log.info("{}", regressionAnalyzer); System.out.println(regressionAnalyzer.toString()); } else { ResultAnalyzer analyzer = new ResultAnalyzer(Arrays.asList(dataset.labels()), "unknown"); for (double[] r : resList) { analyzer.addInstance(dataset.getLabelString(r[0]), new ClassifierResult(dataset.getLabelString(r[1]), 1.0)); } //log.info("{}", analyzer); System.out.println(analyzer.toString()); } }