Package uk.ac.cam.ch.wwmm.oscar3.misc

Examples of uk.ac.cam.ch.wwmm.oscar3.misc.GeniaProcessor

30
31
32
33
34
35
36
37
38
39
40
   * @throws Exception
   */
  public static Document makeModel() throws Exception {
    Element modelRoot = new Element("model");
    modelRoot.appendChild(ExtractTrainingData.getInstance().toXML());
    MEMM memm = MEMMSingleton.getInstance();
    if(memm != null) modelRoot.appendChild(memm.writeModel());
    NESubtypes subtypes = NESubtypes.getInstance();
    if(subtypes.OK) modelRoot.appendChild(subtypes.toXML());
    return new Document(modelRoot);
  }
View Full Code Here
278
279
280
281
282
283
284
285
286
287
288
      }     
    }*/
  }

  void trainOnSbFilesWithRescore(List<File> files, Map<File,String> domains) throws Exception {
    rescorer = new RescoreMEMMOut();
    List<List<File>> splitTrainFiles = new ArrayList<List<File>>();
    List<List<File>> splitTrainAntiFiles = new ArrayList<List<File>>();
    int splitNo = 3;
   
    for(int i=0;i<splitNo;i++) {
View Full Code Here
576
577
578
579
580
581
582
583
584
585
586
        tagSet.add(gm.getOutcome(j));
      }
    }
    Element rescorerElem = memmRoot.getFirstChildElement("rescorer");
    if(rescorerElem != null) {
      rescorer = new RescoreMEMMOut();
      rescorer.readElement(rescorerElem);
    } else {
      rescorer = null;
    }
    makeEntityTypesAndZeroProbs();
View Full Code Here
78
79
80
81
82
83
84
85
86
87
88
    //System.out.println(logAdd(-1000, -1000));
   
    //if(true) return;
   
    NGramBuilder ng = NGramBuilder.getInstance();
    ng.engSet.removeAll(ng.chemSet);
    Map<String,Boolean> isChem = new HashMap<String,Boolean>();
    for(String chem : ng.chemSet) {
      isChem.put(chem, true);
    }
View Full Code Here
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
  }
 
  private boolean resolveVsExternal(ProcessState state, String name) {
    if(extensionNameResolver == null) return false;
    try {
      Results results = extensionNameResolver.resolve(name, state.args);
      if(results != null) {
        state.smiles = results.getSmiles();
        state.inchi = results.getInchi();
        state.cmlMol = results.getCml();
        if(state.smiles != null) setNEAttribute(state.ne, "SMILES", state.smiles);
        if(state.inchi != null) setNEAttribute(state.ne, "InChI", state.inchi);
        if(state.smiles != null && Oscar3Props.getInstance().cacheExtensionNameResolver) {
          cacheEntry(name, state.smiles, state.inchi, state.cmlMol);
        }
View Full Code Here
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
  /**
   * @param args
   */
  public static void main(String[] args) throws Exception {
    ResultDbReader  rdbr = ResultDbReader.getInstance();
    Map<Integer,String> paperIdToPath = new HashMap<Integer, String>();
    Map<String,Document> pathToSaf = new HashMap<String, Document>();
    Map<String,Document> pathToNewSaf = new HashMap<String, Document>();
   
    for(int i=0;i<30000;i++) {
      Document rmrsDoc = rdbr.getRmrs(i);
      if(!rmrsDoc.getRootElement().getLocalName().equals("error")) {
        Serializer ser;
        ser = new Serializer(System.out);
        //ser.setIndent(2);
        ser.write(rmrsDoc);       
        int sentenceId = Integer.parseInt(rmrsDoc.getRootElement().getAttributeValue("sentenceid"));
        Document sentDoc = rdbr.getSentence(sentenceId);
        ser = new Serializer(System.out);
        ser.setIndent(2);
        ser.write(sentDoc);
        String sentContent = sentDoc.query("/sentence/@content").get(0).getValue();
        Nodes neNodes = sentDoc.query("/sentence/ne");
        Map<String,List<String>> neMap = new HashMap<String,List<String>>();
        for (int j=0;j<neNodes.size();j++) {
          Element e = (Element)neNodes.get(j);
          String fromto = e.getAttributeValue("cfrom") + ":" + e.getAttributeValue("cto");
          String safid = e.getAttributeValue("safid").substring(1);
          if(!neMap.containsKey(fromto)) neMap.put(fromto, new ArrayList<String>());
          neMap.get(fromto).add(safid);
        }
        System.out.println(neMap);
        int docId = Integer.parseInt(sentDoc.query("/sentence/@docid").get(0).getValue());
        if(!paperIdToPath.containsKey(docId)) {
          Document paperDoc = rdbr.getPaperInfo(docId);
          paperIdToPath.put(docId, paperDoc.query("/paper/@oscarpath").get(0).getValue());
        }
        if(!pathToSaf.containsKey(paperIdToPath.get(docId))) {
          pathToSaf.put(paperIdToPath.get(docId), new Builder().build(new File(paperIdToPath.get(docId), "saf.xml")));
          File newSaf = new File(new File("/usr/groups/sciborg/texts/ne_annotated_210409/", new File(paperIdToPath.get(docId)).getName()), "saf.xml");
View Full Code Here
32
33
34
35
36
37
38
39
40
  public static Document makeModel() throws Exception {
    Element modelRoot = new Element("model");
    modelRoot.appendChild(ExtractTrainingData.getInstance().toXML());
    MEMM memm = MEMMSingleton.getInstance();
    if(memm != null) modelRoot.appendChild(memm.writeModel());
    NESubtypes subtypes = NESubtypes.getInstance();
    if(subtypes.OK) modelRoot.appendChild(subtypes.toXML());
    return new Document(modelRoot);
  }
View Full Code Here
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
    long time = System.currentTimeMillis();
    ResourceGetter rg = new ResourceGetter("uk/ac/cam/ch/wwmm/oscar3/chemnamedict/resources/");
   
    Set<String> stops = rg.getStringSet("ChEBIStop.txt");
   
    OBOOntology o = OBOOntology.getInstance();

    Map<String,Set<String>> relatedBySource = new HashMap<String,Set<String>>();
   
    for(String id : o.getTerms().keySet()) {
      if(!id.startsWith("CHEBI:")) continue;
      //if(id.equals("CHEBI:16243")) System.out.println("foo!");
      OntologyTerm term = o.getTerms().get(id);
      Set<String> names = new LinkedHashSet<String>();
      names.add(term.getName());
      String inchi = null;
      String smiles = null;
      //System.out.println(term.getName());
View Full Code Here
148
149
150
151
152
153
154
155
156
157
158
159
160
    }
  }
 
  private Bag<String> expandOntologyBag(Bag<String> ontologyBag) {
    Bag<String> results = new Bag<String>();
    OBOOntology oo = OBOOntology.getInstance();
    for(String id : ontologyBag.getSet()) {
      for(String parent : oo.getIdsForIdWithAncestors(id)) {
        results.add(parent, ontologyBag.getCount(id));
      }
    }
    return results;
  }
View Full Code Here
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
    Map<String,Set<String>> relatedBySource = new HashMap<String,Set<String>>();
   
    for(String id : o.getTerms().keySet()) {
      if(!id.startsWith("CHEBI:")) continue;
      //if(id.equals("CHEBI:16243")) System.out.println("foo!");
      OntologyTerm term = o.getTerms().get(id);
      Set<String> names = new LinkedHashSet<String>();
      names.add(term.getName());
      String inchi = null;
      String smiles = null;
      //System.out.println(term.getName());
      for(Synonym s : term.getSynonyms()){
        //if(id.equals("CHEBI:16243")) System.out.println(s);
        //System.out.println("\t" + s);
        if(s.getType().equals("RELATED InChI")) {
          inchi = s.getSyn();
        } else if(s.getType().equals("RELATED SMILES")) {
View Full Code Here
TOP

Related Classes of uk.ac.cam.ch.wwmm.oscar3.misc.GeniaProcessor

  • dk.brics.automaton.Automaton
  • dk.brics.automaton.RunAutomaton
  • jregex.Pattern
  • nu.xom.Attribute
  • nu.xom.Builder
  • nu.xom.Document
  • nu.xom.Element
  • nu.xom.Elements
  • nu.xom.Node
  • nu.xom.Nodes
    • Copyright © 2018 www.massapicom. 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.