in opennlp-tools/src/main/java/opennlp/tools/cmdline/namefind/TokenNameFinderCrossValidatorTool.java [68:163]
public void run(String format, String[] args) {
super.run(format, args);
mlParams = CmdLineUtil.loadTrainingParameters(params.getParams(), true);
if (mlParams == null) {
mlParams = new TrainingParameters();
}
byte[] featureGeneratorBytes =
TokenNameFinderTrainerTool.openFeatureGeneratorBytes(params.getFeaturegen());
Map<String, Object> resources;
try {
resources = TokenNameFinderTrainerTool.loadResources(params.getResources(), params.getFeaturegen());
}
catch (IOException e) {
throw new TerminateToolException(-1,"IO error while loading resources", e);
}
if (params.getNameTypes() != null) {
String[] nameTypes = params.getNameTypes().split(",");
sampleStream = new NameSampleTypeFilter(nameTypes, sampleStream);
}
List<EvaluationMonitor<NameSample>> listeners = new LinkedList<>();
if (params.getMisclassified()) {
listeners.add(new NameEvaluationErrorListener());
}
String sequenceCodecImplName = params.getSequenceCodec();
if ("BIO".equals(sequenceCodecImplName)) {
sequenceCodecImplName = BioCodec.class.getName();
}
else if ("BILOU".equals(sequenceCodecImplName)) {
sequenceCodecImplName = BilouCodec.class.getName();
}
SequenceCodec<String> sequenceCodec;
try {
sequenceCodec = TokenNameFinderFactory.instantiateSequenceCodec(sequenceCodecImplName);
} catch (InvalidFormatException e) {
throw new TerminateToolException(-1, e.getMessage(), e);
}
TokenNameFinderFineGrainedReportListener reportListener = null;
File reportFile = params.getReportOutputFile();
OutputStream reportOutputStream = null;
if (reportFile != null) {
CmdLineUtil.checkOutputFile("Report Output File", reportFile);
try {
reportOutputStream = new FileOutputStream(reportFile);
reportListener = new TokenNameFinderFineGrainedReportListener(sequenceCodec,
reportOutputStream);
listeners.add(reportListener);
} catch (FileNotFoundException e) {
throw new TerminateToolException(-1,
"IO error while creating Name Finder fine-grained report file: "
+ e.getMessage());
}
}
TokenNameFinderFactory nameFinderFactory;
try {
nameFinderFactory = TokenNameFinderFactory.create(params.getFactory(),
featureGeneratorBytes, resources, sequenceCodec);
} catch (InvalidFormatException e) {
throw new TerminateToolException(-1, e.getMessage(), e);
}
TokenNameFinderCrossValidator validator;
try {
validator = new TokenNameFinderCrossValidator(params.getLang(),
params.getType(), mlParams, nameFinderFactory,
listeners.toArray(new TokenNameFinderEvaluationMonitor[0]));
validator.evaluate(sampleStream, params.getFolds());
} catch (IOException e) {
throw createTerminationIOException(e);
} finally {
try {
sampleStream.close();
} catch (IOException e) {
// sorry that this can fail
}
}
logger.info("done");
if (reportFile != null) {
reportListener.writeReport();
}
logger.info(validator.getFMeasure().toString());
}