/**
* Make Document for coref (for method coref(Document doc, StringBuilder[] outputs)). Mention
* detection and document preprocessing is done here.
*
* @throws Exception
*/
public Document makeDocument(InputDoc input) throws Exception {
if (input == null) return null;
Annotation anno = input.annotation;
// add missing annotation
if (needMissingAnnotations) {
addMissingAnnotation(anno);
}
if (Boolean.parseBoolean(props.getProperty("hcoref.useMarkedDiscourse", "false"))) {
anno.set(CoreAnnotations.UseMarkedDiscourseAnnotation.class, true);
}
// remove nested NP with same headword except newswire document for chinese
if (input.conllDoc != null && CorefProperties.getLanguage(props) == Locale.CHINESE) {
CorefProperties.setRemoveNested(props, !input.conllDoc.documentID.contains("nw"));
}
// mention detection: MD gives following information about mentions: mention start/end index,
// span, headword
// rest information will be set in preprocess step
List<List<Mention>> mentions = md.findMentions(anno, dict, props);
Document doc = new Document(input, mentions);
// find headword for gold mentions
if (input.goldMentions != null) findGoldMentionHeads(doc);
// document preprocessing: initialization (assign ID), mention processing (gender, number, type,
// etc), speaker extraction, etc
Preprocessor.preprocess(doc, dict, singletonPredictor, headFinder);
return doc;
}
public static void main(String[] argv)
throws IOException, CmdLineParser.UnknownOptionException,
CmdLineParser.IllegalOptionValueException {
if (argv.length < 4) {
System.out.println("Usage: hictools pre <options> <inputFile> <outputFile> <genomeID>");
System.out.println(" <options>: -d only calculate intra chromosome (diagonal) [false]");
System.out.println(
" : -o calculate densities (observed/expected), write to file [false]");
System.out.println(" : -t <int> only write cells with count above threshold t [0]");
System.out.println(
" : -c <chromosome ID> only calculate map on specific chromosome");
System.exit(0);
}
Globals.setHeadless(true);
CommandLineParser parser = new CommandLineParser();
parser.parse(argv);
String[] args = parser.getRemainingArgs();
if (args[0].equals("sort")) {
AlignmentsSorter.sort(args[1], args[2], null);
} else if (args[0].equals("pairsToBin")) {
String ifile = args[1];
String ofile = args[2];
String genomeId = args[3];
List<Chromosome> chromosomes = loadChromosomes(genomeId);
AsciiToBinConverter.convert(ifile, ofile, chromosomes);
} else if (args[0].equals("binToPairs")) {
String ifile = args[1];
String ofile = args[2];
AsciiToBinConverter.convertBack(ifile, ofile);
} else if (args[0].equals("printmatrix")) {
if (args.length < 5) {
System.err.println(
"Usage: hictools printmatrix <observed/oe/pearson> hicFile chr1 chr2 binsize");
System.exit(-1);
}
String type = args[1];
String file = args[2];
String chr1 = args[3];
String chr2 = args[4];
String binSizeSt = args[5];
int binSize = 0;
try {
binSize = Integer.parseInt(binSizeSt);
} catch (NumberFormatException e) {
System.err.println("Integer expected. Found: " + binSizeSt);
System.exit(-1);
}
dumpMatrix(file, chr1, chr2, binSize, type);
} else if (args[0].equals("eigenvector")) {
if (args.length < 4) {
System.err.println("Usage: hictools eigenvector hicFile chr binsize");
}
String file = args[1];
String chr = args[2];
String binSizeSt = args[3];
int binSize = 0;
try {
binSize = Integer.parseInt(binSizeSt);
} catch (NumberFormatException e) {
System.err.println("Integer expected. Found: " + binSizeSt);
System.exit(-1);
}
calculateEigenvector(file, chr, binSize);
} else if (args[0].equals("pre")) {
String genomeId = "";
try {
genomeId = args[3];
} catch (ArrayIndexOutOfBoundsException e) {
System.err.println("No genome ID given");
System.exit(0);
}
List<Chromosome> chromosomes = loadChromosomes(genomeId);
long genomeLength = 0;
for (Chromosome c : chromosomes) {
if (c != null) genomeLength += c.getSize();
}
chromosomes.set(0, new Chromosome(0, "All", (int) (genomeLength / 1000)));
String[] tokens = args[1].split(",");
List<String> files = new ArrayList<String>(tokens.length);
for (String f : tokens) {
files.add(f);
}
Preprocessor preprocessor = new Preprocessor(new File(args[2]), chromosomes);
preprocessor.setIncludedChromosomes(parser.getChromosomeOption());
preprocessor.setCountThreshold(parser.getCountThresholdOption());
preprocessor.setNumberOfThreads(parser.getThreadedOption());
preprocessor.setDiagonalsOnly(parser.getDiagonalsOption());
preprocessor.setLoadDensities(parser.getDensitiesOption());
preprocessor.preprocess(files);
}
}
public void toTDF(
String typeString,
String ifile,
String ofile,
String probeFile,
String genomeId,
int maxZoomValue,
Collection<WindowFunction> windowFunctions,
String tmpDirName,
int maxRecords)
throws IOException, PreprocessingException {
if (!ifile.endsWith(".affective.csv")) validateIsTilable(typeString);
System.out.println("toTDF. File = " + ifile);
System.out.println("Max zoom = " + maxZoomValue);
if (probeFile != null && probeFile.trim().length() > 0) {
System.out.println("Probe file = " + probeFile);
}
System.out.print("Window functions: ");
for (WindowFunction wf : windowFunctions) {
System.out.print(wf.toString() + " ");
}
System.out.println();
boolean isGCT = isGCT(typeString);
Genome genome = loadGenome(genomeId, isGCT);
if (genome == null) {
throw new PreprocessingException("Genome could not be loaded: " + genomeId);
}
File inputFileOrDir = new File(ifile);
// Estimae the total number of lines to be parsed, for progress updates
int nLines = estimateLineCount(inputFileOrDir);
// TODO -- move this block of code out of here, this should be done before calling this method
// Convert gct files to igv format first
File deleteme = null;
if (isGCT(typeString)) {
File tmpDir = null;
if (tmpDirName != null && tmpDirName.length() > 0) {
tmpDir = new File(tmpDirName);
if (!tmpDir.exists() || !tmpDir.isDirectory()) {
throw new PreprocessingException(
"Specified tmp directory does not exist or is not directory: " + tmpDirName);
}
} else {
tmpDir = new File(System.getProperty("java.io.tmpdir"), System.getProperty("user.name"));
}
if (!tmpDir.exists()) {
tmpDir.mkdir();
}
String baseName = (new File(ifile)).getName();
File igvFile = new File(tmpDir, baseName + ".igv");
igvFile.deleteOnExit();
doGCTtoIGV(typeString, ifile, igvFile, probeFile, maxRecords, tmpDirName, genome);
inputFileOrDir = igvFile;
deleteme = igvFile;
typeString = ".igv";
}
// Convert to tdf
File outputFile = new File(ofile);
try {
Preprocessor p = new Preprocessor(outputFile, genome, windowFunctions, nLines, null);
if (inputFileOrDir.isDirectory() || inputFileOrDir.getName().endsWith(".list")) {
List<File> files = getFilesFromDirOrList(inputFileOrDir);
for (File f : files) {
p.preprocess(f, maxZoomValue, typeString);
}
} else {
p.preprocess(inputFileOrDir, maxZoomValue, typeString);
}
p.finish();
} catch (IOException e) {
e.printStackTrace();
// Delete output file as its probably corrupt
if (outputFile.exists()) {
outputFile.delete();
}
} finally {
if (deleteme != null && deleteme.exists()) {
deleteme.delete();
}
}
System.out.flush();
}
package antenna.preprocessor.v3;