protected boolean readTAssignFile(String tassignFile) {
try {
int i, j;
BufferedReader reader =
new BufferedReader(new InputStreamReader(new FileInputStream(tassignFile), "UTF-8"));
String line;
// z = new Vector[M];
// z = new ArrayList[M];
z = new short[M][];
data = new LDADataset(M);
data.V = V;
for (i = 0; i < M; i++) {
line = reader.readLine();
StringTokenizer tknr = new StringTokenizer(line, " \t\r\n");
int length = tknr.countTokens();
// Vector<Integer> words = new Vector<Integer>();
// Vector<Integer> topics = new Vector<Integer>();
ArrayList<Integer> words = new ArrayList<>();
ArrayList<Integer> topics = new ArrayList<>();
for (j = 0; j < length; j++) {
String token = tknr.nextToken();
StringTokenizer tknr2 = new StringTokenizer(token, ":");
if (tknr2.countTokens() != 2) {
System.out.println("Invalid word-topic assignment line\n");
return false;
}
words.add(Integer.parseInt(tknr2.nextToken()));
topics.add(Integer.parseInt(tknr2.nextToken()));
} // end for each topic assignment
// allocate and add new document to the corpus
Document doc = new Document(words);
data.setDoc(doc, i);
// assign values for z
// z[i] = new Vector<Integer>();
// z[i] = new ArrayList<>(topics.size()); // Note: Modify: init arraylist
// size.
z[i] = new short[topics.size()];
for (j = 0; j < topics.size(); j++) {
// z[i].add(topics.get(j));
z[i][j] = (short) (int) topics.get(j);
}
} // end for each doc
reader.close();
} catch (Exception e) {
System.out.println("Error while loading model: " + e.getMessage());
e.printStackTrace();
return false;
}
return true;
}
