Package cc.mallet.grmm.learning.extract

Source Code of cc.mallet.grmm.learning.extract.AcrfExtractorTui

/* Copyright (C) 2003 Univ. of Massachusetts Amherst, Computer Science Dept.
   This file is part of "MALLET" (MAchine Learning for LanguagE Toolkit).
   http://www.cs.umass.edu/~mccallum/mallet
   This software is provided under the terms of the Common Public License,
   version 1.0, as published by http://www.opensource.org.  For further
   information, see the file `LICENSE' included with this distribution. */
package cc.mallet.grmm.learning.extract;
/**
*
* Created: Aug 23, 2005
*
* @author <A HREF="mailto:casutton@cs.umass.edu>casutton@cs.umass.edu</A>
* @version $Id: AcrfExtractorTui.java,v 1.1 2007/10/22 21:38:02 mccallum Exp $
*/

import bsh.EvalError;

import java.io.BufferedReader;
import java.io.File;
import java.io.FileReader;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Iterator;
import java.util.LinkedList;
import java.util.List;
import java.util.logging.Logger;
import java.util.regex.Pattern;

import cc.mallet.extract.Extraction;
import cc.mallet.extract.ExtractionEvaluator;
import cc.mallet.grmm.inference.Inferencer;
import cc.mallet.grmm.learning.*;
import cc.mallet.pipe.*;
import cc.mallet.pipe.iterator.FileListIterator;
import cc.mallet.pipe.iterator.LineGroupIterator;
import cc.mallet.pipe.iterator.PipeInputIterator;
import cc.mallet.types.Instance;
import cc.mallet.types.InstanceList;
import cc.mallet.util.*;

public class AcrfExtractorTui {

  private static final Logger logger = MalletLogger.getLogger (AcrfExtractorTui.class.getName ());

  private static CommandOption.File outputPrefix = new CommandOption.File
          (AcrfExtractorTui.class, "output-prefix", "FILENAME", true, null,
             "Directory to write saved model to.", null);

  private static CommandOption.File modelFile = new CommandOption.File
          (AcrfExtractorTui.class, "model-file", "FILENAME", true, null, "Text file describing model structure.", null);

  private static CommandOption.File trainFile = new CommandOption.File
          (AcrfExtractorTui.class, "training", "FILENAME", true, null, "File containing training data.", null);

  private static CommandOption.File testFile = new CommandOption.File
          (AcrfExtractorTui.class, "testing", "FILENAME", true, null, "File containing testing data.", null);

  private static CommandOption.Integer numLabelsOption = new CommandOption.Integer
  (AcrfExtractorTui.class, "num-labels", "INT", true, -1,
          "If supplied, number of labels on each line of input file." +
                  "  Otherwise, the token ---- must separate labels from features.", null);

  private static CommandOption.String trainerOption = new CommandOption.String
          (AcrfExtractorTui.class, "trainer", "STRING", true, "ACRFExtractorTrainer",
                  "Specification of trainer type.", null);

  private static CommandOption.String inferencerOption = new CommandOption.String
          (AcrfExtractorTui.class, "inferencer", "STRING", true, "LoopyBP",
                  "Specification of inferencer.", null);

  private static CommandOption.String maxInferencerOption = new CommandOption.String
          (AcrfExtractorTui.class, "max-inferencer", "STRING", true, "LoopyBP.createForMaxProduct()",
                  "Specification of inferencer.", null);

  private static CommandOption.String evalOption = new CommandOption.String
          (AcrfExtractorTui.class, "eval", "STRING", true, "LOG",
                  "Evaluator to use.  Java code grokking performed.", null);

  private static CommandOption.String extractionEvalOption = new CommandOption.String
          (AcrfExtractorTui.class, "extraction-eval", "STRING", true, "PerDocumentF1",
                  "Evaluator to use.  Java code grokking performed.", null);

  private static CommandOption.Integer checkpointIterations = new CommandOption.Integer
    (AcrfExtractorTui.class, "checkpoint", "INT", true, -1, "Save a copy after every ___ iterations.", null);


  static CommandOption.Boolean cacheUnrolledGraph = new CommandOption.Boolean
          (AcrfExtractorTui.class, "cache-graphs", "true|false", true, true,
                  "Whether to use memory-intensive caching.", null);

  static CommandOption.Boolean perTemplateTrain = new CommandOption.Boolean
          (AcrfExtractorTui.class, "per-template-train", "true|false", true, false,
                  "Whether to pretrain templates before joint training.", null);

  static CommandOption.Integer pttIterations = new CommandOption.Integer
          (AcrfExtractorTui.class, "per-template-iterations", "INTEGER", false, 100,
                  "How many training iterations for each step of per-template-training.", null);

  static CommandOption.Integer randomSeedOption = new CommandOption.Integer
  (AcrfExtractorTui.class, "random-seed", "INTEGER", true, 0,
   "The random seed for randomly selecting a proportion of the instance list for training", null);

  static CommandOption.Boolean useTokenText = new CommandOption.Boolean
          (AcrfExtractorTui.class, "use-token-text", "true|false", true, true,
           "If true, first feature in list is assumed to be token identity, and is treated specially.", null);
 
  private static CommandOption.Boolean labelsAtEnd = new CommandOption.Boolean
  (AcrfExtractorTui.class, "labels-at-end", "INT", true, false,
   "If true, then label is at end of each line, rather than beginning.", null);

  static CommandOption.Boolean trainingIsList = new CommandOption.Boolean
          (AcrfExtractorTui.class, "training-is-list", "true|false", true, false,
           "If true, training option gives list of files to read for training.", null);

  private static CommandOption.File dataDir = new CommandOption.File
          (AcrfExtractorTui.class, "data-dir", "FILENAME", true, null, "If training-is-list, base directory in which training files located.", null);


  private static BshInterpreter interpreter = setupInterpreter ();

  public static void main (String[] args) throws IOException, EvalError
  {
    doProcessOptions (AcrfExtractorTui.class, args);
    Timing timing = new Timing ();

    GenericAcrfData2TokenSequence basePipe;
    if (!numLabelsOption.wasInvoked ()) {
      basePipe = new GenericAcrfData2TokenSequence ();
    } else {
      basePipe = new GenericAcrfData2TokenSequence (numLabelsOption.value);
    }

    if (!useTokenText.value) {
      basePipe.setFeaturesIncludeToken(false);
      basePipe.setIncludeTokenText(false);
    }

    basePipe.setLabelsAtEnd (labelsAtEnd.value);

    Pipe tokPipe = new SerialPipes (new Pipe[] {
        (trainingIsList.value ? new Input2CharSequence () : (Pipe) new Noop ()),
        basePipe,
    });

    Iterator<Instance> trainSource = constructIterator(trainFile.value, dataDir.value, trainingIsList.value);
    Iterator<Instance> testSource;
    if (testFile.wasInvoked ()) {
      testSource = constructIterator (testFile.value, dataDir.value, trainingIsList.value);
    } else {
      testSource = null;
    }

    ACRF.Template[] tmpls = parseModelFile (modelFile.value);

    ACRFExtractorTrainer trainer = createTrainer (trainerOption.value);
    ACRFEvaluator eval = createEvaluator (evalOption.value);
    ExtractionEvaluator extractionEval = createExtractionEvaluator (extractionEvalOption.value);

    Inferencer inf = createInferencer (inferencerOption.value);
    Inferencer maxInf = createInferencer (maxInferencerOption.value);

    trainer.setPipes (tokPipe, new TokenSequence2FeatureVectorSequence ())
            .setDataSource (trainSource, testSource)
            .setEvaluator (eval)
            .setTemplates (tmpls)
            .setInferencer (inf)
            .setViterbiInferencer (maxInf)
            .setCheckpointDirectory (outputPrefix.value)
            .setNumCheckpointIterations (checkpointIterations.value)
            .setCacheUnrolledGraphs (cacheUnrolledGraph.value)
            .setUsePerTemplateTrain (perTemplateTrain.value)
            .setPerTemplateIterations (pttIterations.value);

    logger.info ("Starting training...");
    ACRFExtractor extor = trainer.trainExtractor ();
    timing.tick ("Training");

    FileUtils.writeGzippedObject (new File (outputPrefix.value, "extor.ser.gz"), extor);
    timing.tick ("Serializing");

    InstanceList testing = trainer.getTestingData ();
    if (testing != null) {
      eval.test (extor.getAcrf (), testing, "Final results");
    }

    if ((extractionEval != null) && (testing != null)) {
      Extraction extraction = extor.extract (testing);
      extractionEval.evaluate (extraction);
      timing.tick ("Evaluting");
    }

    System.out.println ("Total time (ms) = " + timing.elapsedTime ());
  }

  private static BshInterpreter setupInterpreter ()
  {
    BshInterpreter interpreter = CommandOption.getInterpreter ();
    try {
      interpreter.eval ("import edu.umass.cs.mallet.base.extract.*");
      interpreter.eval ("import edu.umass.cs.mallet.grmm.inference.*");
      interpreter.eval ("import edu.umass.cs.mallet.grmm.learning.*");
      interpreter.eval ("import edu.umass.cs.mallet.grmm.learning.templates.*");
      interpreter.eval ("import edu.umass.cs.mallet.grmm.learning.extract.*");
    } catch (EvalError e) {
      throw new RuntimeException (e);
    }

    return interpreter;
  }


  private static Iterator<Instance> constructIterator (File trainFile, File dataDir, boolean isList) throws IOException
  {
    if (isList) {
      return new FileListIterator (trainFile, dataDir, null, null, true);
    } else {
      return new LineGroupIterator (new FileReader (trainFile), Pattern.compile ("^\\s*$"), true);
    }
  }

  public static ACRFEvaluator createEvaluator (String spec) throws EvalError
  {
    if (spec.indexOf ('(') >= 0) {
      // assume it's Java code, and don't screw with it.
      return (ACRFEvaluator) interpreter.eval (spec);
    } else {
      LinkedList toks = new LinkedList (Arrays.asList (spec.split ("\\s+")));
      return createEvaluator (toks);
    }
  }

  private static ExtractionEvaluator createExtractionEvaluator (String spec) throws EvalError
  {
    if (spec.indexOf ('(') >= 0) {
      // assume it's Java code, and don't screw with it.
      return (ExtractionEvaluator) interpreter.eval (spec);
    } else {
      spec = "new "+spec+"Evaluator ()";
      return (ExtractionEvaluator) interpreter.eval (spec);
    }
  }

  private static ACRFEvaluator createEvaluator (LinkedList toks)
  {
    String type = (String) toks.removeFirst ();

    if (type.equalsIgnoreCase ("SEGMENT")) {
      int slice = Integer.parseInt ((String) toks.removeFirst ());
      if (toks.size() % 2 != 0)
        throw new RuntimeException ("Error in --eval "+evalOption.value+": Every start tag must have a continue.");
      int numTags = toks.size () / 2;
      String[] startTags = new String [numTags];
      String[] continueTags = new String [numTags];

      for (int i = 0; i < numTags; i++) {
        startTags[i] = (String) toks.removeFirst ();
        continueTags[i] = (String) toks.removeFirst ();
      }

      return new MultiSegmentationEvaluatorACRF (startTags, continueTags, slice);

    } else if (type.equalsIgnoreCase ("LOG")) {
      return new DefaultAcrfTrainer.LogEvaluator ();

    } else if (type.equalsIgnoreCase ("SERIAL")) {
      List evals = new ArrayList ();
      while (!toks.isEmpty ()) {
        evals.add (createEvaluator (toks));
      }
      return new AcrfSerialEvaluator (evals);

    } else {
      throw new RuntimeException ("Error in --eval "+evalOption.value+": illegal evaluator "+type);
    }
  }

  private static ACRFExtractorTrainer createTrainer (String spec) throws EvalError
  {
    String cmd;
    if (spec.indexOf ('(') >= 0) {
      // assume it's Java code, and don't screw with it.
      cmd = spec;
    } else if (spec.endsWith ("Trainer")) {
      cmd = "new "+spec+"()";
    } else {
      cmd = "new "+spec+"Trainer()";
    }

    // Return whatever the Java code says to
    Object trainer = interpreter.eval (cmd);

    if (trainer instanceof ACRFExtractorTrainer)
      return (ACRFExtractorTrainer) trainer;
    else if (trainer instanceof DefaultAcrfTrainer)
      return new ACRFExtractorTrainer ().setTrainingMethod ((ACRFTrainer) trainer);
    else throw new RuntimeException ("Don't know what to do with trainer "+trainer);
  }

  private static Inferencer createInferencer (String spec) throws EvalError
  {
    String cmd;
    if (spec.indexOf ('(') >= 0) {
      // assume it's Java code, and don't screw with it.
      cmd = spec;
    } else {
      cmd = "new "+spec+"()";
    }

    // Return whatever the Java code says to
    Object inf = interpreter.eval (cmd);

    if (inf instanceof Inferencer)
      return (Inferencer) inf;

    else throw new RuntimeException ("Don't know what to do with inferencer "+inf);
  }


  public static void doProcessOptions (Class childClass, String[] args)
  {
    CommandOption.List options = new CommandOption.List ("", new CommandOption[0]);
    options.add (childClass);
    options.process (args);
    options.logOptions (Logger.getLogger (""));
  }

  private static ACRF.Template[] parseModelFile (File mdlFile) throws IOException, EvalError
  {
    BufferedReader in = new BufferedReader (new FileReader (mdlFile));

    List tmpls = new ArrayList ();
    String line = in.readLine ();
    while (line != null) {
      Object tmpl = interpreter.eval (line);
      if (!(tmpl instanceof ACRF.Template)) {
        throw new RuntimeException ("Error in "+mdlFile+" line "+in.toString ()+":\n  Object "+tmpl+" not a template");
      }
      tmpls.add (tmpl);
      line = in.readLine ();
    }

    return (ACRF.Template[]) tmpls.toArray (new ACRF.Template [0]);
  }

}
TOP

Related Classes of cc.mallet.grmm.learning.extract.AcrfExtractorTui

TOP
Copyright © 2018 www.massapi.com. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.