Package ivory.core.data.dictionary

Examples of ivory.core.data.dictionary.DefaultCachedFrequencySortedDictionary

153
154
155
156
157
158
159
160
161
162
163
    postingsIndex = new IntPostingsForwardIndex(indexPath, fs);
    LOG.info(" - Number of terms: " + readCollectionTermCount());
    LOG.info("Done!");

    try {
      termidMap = new DefaultCachedFrequencySortedDictionary(new Path(getIndexTermsData()), new Path(getIndexTermIdsData()),
          new Path(getIndexTermIdMappingData()), 0.2f,  fs);
    } catch (Exception e) {
      throw new ConfigurationException("Error initializing term to term id mapping!", e);
    }
View Full Code Here
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
        String termidsFile = env.getIndexTermIdsData();
        String idToTermFile = env.getIndexTermIdMappingData();

        // Take a different code path if we're in standalone mode.
        if (conf.get("mapred.job.tracker").equals("local")) {
          dictionary = new DefaultCachedFrequencySortedDictionary(new Path(termsFile),
              new Path(termidsFile), new Path(idToTermFile), 0.3f, FileSystem.getLocal(conf));
        } else {
          // We need to figure out which file in the DistributeCache is which...
          Map<String, Path> pathMapping = Maps.newHashMap();
          Path[] localFiles = DistributedCache.getLocalCacheFiles(context.getConfiguration());
          for (Path p : localFiles) {
            LOG.info("In DistributedCache: " + p);
            if (p.toString().contains(termsFile)) {
              pathMapping.put(termsFile, p);
            } else if (p.toString().contains(termidsFile)) {
              pathMapping.put(termidsFile, p);
            } else if (p.toString().contains(idToTermFile)) {
              pathMapping.put(idToTermFile, p);
            }
          }

          LOG.info(" - terms: " + pathMapping.get(termsFile));
          LOG.info(" - id: " + pathMapping.get(termidsFile));
          LOG.info(" - idToTerms: " + pathMapping.get(idToTermFile));

          dictionary = new DefaultCachedFrequencySortedDictionary(pathMapping.get(termsFile),
              pathMapping.get(termidsFile), pathMapping.get(idToTermFile),
              0.3f, FileSystem.getLocal(context.getConfiguration()));
        }
      } catch (Exception e) {
        e.printStackTrace();
View Full Code Here
59
60
61
62
63
64
65
66
67
68
69
        gs.loadDFStats(localFiles[1], fs);
        gs.loadCFStats(localFiles[2], fs);

        String indexPath = job.get("Ivory.IndexPath");
        sLogger.info("loading TermIdMap from " + indexPath);
        mTermIdMap = new DefaultCachedFrequencySortedDictionary(localFiles[3], localFiles[4],
            localFiles[5], 0.2f, fs);
      } catch (IOException e) {
        e.printStackTrace();
        throw new RuntimeException("Error loading global term stats!");
      }
View Full Code Here
TOP

Related Classes of ivory.core.data.dictionary.DefaultCachedFrequencySortedDictionary

  • ivory.core.index.DistributeGlobalStatsToPostings$MyMapper
  • ivory.core.preprocess.BuildIntDocVectors$MyMapper
  • ivory.core.RetrievalEnvironment
  • org.apache.hadoop.conf.Configuration
  • org.apache.hadoop.fs.FileSystem
  • org.apache.hadoop.fs.Path
  • java.io.InputStreamReader
  • java.io.BufferedReader
    • 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.