Package com.googlecode.gaal.data.api

Examples of com.googlecode.gaal.data.api.IntSequence

302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
        }
        return repeatsMap;
    }

    private static List<String> toDocumentList(Corpus<String> corpus, Map<Integer, Integer> repeatsMap) {
        IntSequence sequence = corpus.sequence();
        List<String> documents = new ArrayList<String>();
        StringBuilder sb = new StringBuilder();
        int prevRepeatNumber = -1;
        int docNumber = 0;
        boolean isFirst = true;
        boolean isBlockStart = false;
        for (int i = 0; i < sequence.size(); i++) {
            Integer repeatNumber = repeatsMap.get(i);
            if (prevRepeatNumber != -1 && (repeatNumber == null || prevRepeatNumber != repeatNumber)) {
                sb.append("}");
            }
            if (repeatNumber != null && prevRepeatNumber != repeatNumber) {
                sb.append(String.format(" {\\color{%s}", TikzConstants.VECTOR_STYLES[repeatNumber].getColour()));
                isBlockStart = true;
            }
            if (repeatNumber != null) {
                prevRepeatNumber = repeatNumber;
            } else {
                prevRepeatNumber = -1;
            }
            int docId = corpus.getDocumentId(i);
            if (docNumber != docId) {
                docNumber = docId;
                documents.add(sb.toString());
                sb = new StringBuilder();
            }
            int symbol = sequence.get(i);
            String token = corpus.toToken(symbol);
            if (!corpus.isSeparator(symbol)) {
                if (isFirst) {
                    isFirst = false;
                } else if (!token.equals(",") && !token.equals(".") && !isBlockStart) {
View Full Code Here
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
        documents.add(sb.toString());
        return documents;
    }

    private static String toDocument(Corpus<String> corpus) {
        IntSequence sequence = corpus.sequence();
        StringBuilder sb = new StringBuilder();
        boolean isFirst = true;
        for (int i = 0; i < sequence.size(); i++) {
            int symbol = sequence.get(i);
            String token = corpus.toToken(symbol);
            if (!corpus.isSeparator(symbol)) {
                if (isFirst) {
                    isFirst = false;
                } else if (!token.equals(",") && !token.equals(".")) {
View Full Code Here
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
        return sb.toString();
    }

    private static List<Map<String, Integer>> toTermMapList(Corpus<String> corpus) {
        List<Map<String, Integer>> documentList = new ArrayList<Map<String, Integer>>();
        IntSequence sequence = corpus.sequence();
        Map<String, Integer> termMap = null;
        int docNumber = -1;
        boolean isFirst = true;
        for (int i = 0; i < sequence.size(); i++) {
            int docId = corpus.getDocumentId(i);
            if (docNumber != docId) {
                docNumber = docId;
                if (termMap != null)
                    documentList.add(termMap);
                termMap = new TreeMap<String, Integer>();
            }
            int symbol = sequence.get(i);
            String token = corpus.toToken(symbol);
            if (!corpus.isSeparator(symbol)) {
                if (isFirst) {
                    isFirst = false;
                } else if (!token.equals(",") && !token.equals(".")) {
View Full Code Here
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
    }

    private <E extends BinaryNode<E>> int traverse(IntervalSet<E> bwtSet, SuffixArray sa, E interval) {
        if (interval.isTerminal()) {
            int loc = sa.getSuffixTable()[interval.left()];
            IntSequence sequence = sa.getSequence();
            if (loc == 0)
                return sequence.get(sequence.size() - 1);
            else
                return sequence.get(loc - 1);
        } else {
            E left = interval.leftChild();
            E right = interval.rightChild();

            int leftValue = traverse(bwtSet, sa, left);
View Full Code Here
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
     * @return the embedded suffix tree for the given interval, null if such
     *         tree can not be constructed
     */
    public static <S> EmbeddedSuffixTree create(SuffixArray sa, Interval interval, int windowSize,
            SymbolTable<S> symbolTable) {
        IntSequence sequence = sa.getSequence();
        int[] suffixTable = sa.getSuffixTable();
        int[] inverseSuffixTable = sa.getInverseSuffixTable();
        SortedMap<Integer, Integer> embeddedSuffixTableIndices = new TreeMap<Integer, Integer>();
        int lcp = interval.lcp();
        IntSequence indices = interval.indices();
        for (int i = 0; i < interval.size(); i++) {
            int start = indices.get(i) + lcp;
            for (int j = start; j < start + windowSize && j < inverseSuffixTable.length
                    && !symbolTable.isSeparator(sequence.get(j)); j++) {
                Integer startIndex = embeddedSuffixTableIndices.get(inverseSuffixTable[j]);
                if (startIndex == null || startIndex < start) {
                    embeddedSuffixTableIndices.put(inverseSuffixTable[j], start);
View Full Code Here
57
58
59
60
61
62
63
64
65
66
67
    public String toString(Iterator<IntSequence> iterator, String delimiter) {
        StringBuilder buffer = new StringBuilder();
        boolean isFirst = true;
        buffer.append('{');
        while (iterator.hasNext()) {
            IntSequence sequence = iterator.next();
            if (isFirst)
                isFirst = false;
            else
                buffer.append(',');
            buffer.append(String.format("\"%s\"", join(iterator(sequence), delimiter)));
View Full Code Here
25
26
27
28
29
30
31
32
33
34
35
36
37
        }
    }

    public static <S> Vector toVector(Interval interval, Corpus<S> corpus) {
        Vector vector = new SparseVector();
        IntSequence indices = interval.indices();
        for (int i = 0; i < indices.size(); i++) {
            int start = indices.get(i);
            int documentId = corpus.getDocumentId(start);
            vector.add(documentId, 1);
        }
        return vector;
    }
View Full Code Here
19
20
21
22
23
24
25
26
            return true;
        if (obj == null)
            return false;
        if (getClass() != obj.getClass())
            return false;
        IntSequence other = (IntSequence) obj;
        return compareTo(other) == 0;
    }
View Full Code Here
135
136
137
138
139
140
141
142
143
144
145
146
147
        };
    }

    public static Vector toVector(Interval interval, Corpus<String> corpus) {
        Vector vector = new SparseVector();
        IntSequence indices = interval.indices();
        for (int i = 0; i < indices.size(); i++) {
            int start = indices.get(i);
            int documentId = corpus.getDocumentId(start);
            vector.add(documentId, 1);
        }
        return vector;
    }
View Full Code Here
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
        int prevSuffix = -1;
        boolean hasCommonSuffix = true; // all the fillers share a common
                                        // suffix
        Iterator<IntSequence> fillerIterator = embeddedInterval.fillerIterator();
        while (fillerIterator.hasNext()) {
            IntSequence filler =  fillerIterator.next();
            if (prevPrefix != -1 && hasCommonPrefix && (filler.size() == 0 || filler.get(0) != prevPrefix)) {
                hasCommonPrefix = false;
            }
            if (prevSuffix != -1 && hasCommonSuffix
                    && (filler.size() == 0 || filler.get(filler.size() - 1) != prevSuffix)) {
                hasCommonSuffix = false;
            }
            if (filler.size() > 0) {
                prevPrefix = filler.get(0);
                prevSuffix = filler.get(filler.size() - 1);
                nonEmptyFillerCount++;
            }
        }
        return (!hasCommonPrefix && !hasCommonSuffix && leftRightContextRatio > minLeftRightContextRatio && nonEmptyFillerCount > minNonEmptyFillerCount);
    }
View Full Code Here
TOP

Related Classes of com.googlecode.gaal.data.api.IntSequence

  • com.googlecode.gaal.analysis.impl.Aligner
  • com.googlecode.gaal.analysis.impl.ContextExtractorTest
  • com.googlecode.gaal.analysis.impl.IntervalVectorBuilder
  • com.googlecode.gaal.analysis.impl.NaiveDistinctBwtSetBuilder
  • com.googlecode.gaal.analysis.impl.NonSingletonBwtSetBuilder
  • com.googlecode.gaal.analysis.impl.SetBuildersTest
  • com.googlecode.gaal.analysis.impl.SimpleContextFilter
  • com.googlecode.gaal.analysis.impl.SingletonBwtSetBuilder
  • com.googlecode.gaal.data.impl.AbstractCorpus
  • com.googlecode.gaal.data.impl.AbstractSequence
    • 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.