本文整理汇总了Java中cc.mallet.topics.ParallelTopicModel.getTopWords方法的典型用法代码示例。如果您正苦于以下问题：Java ParallelTopicModel.getTopWords方法的具体用法？Java ParallelTopicModel.getTopWords怎么用？Java ParallelTopicModel.getTopWords使用的例子？那么, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类cc.mallet.topics.ParallelTopicModel的用法示例。
在下文中一共展示了ParallelTopicModel.getTopWords方法的1个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Java代码示例。
示例1: selectTopLDAFeatures
import cc.mallet.topics.ParallelTopicModel; //导入方法依赖的package包/类
/**
* Select top features in LDA topics.
*
* @param numSelFeatures Number of features to select.
* @param ldaEst LDAEstimatePr which provides an interface to an LDA model.
* @param seqAlphabet The alphabet for the sequence dataset, which may be different from the vector dataset alphabet.
* @param alphabet The vector dataset alphabet.
* @return ArrayList with the int indices of the selected features.
*/
public static ArrayList<Integer> selectTopLDAFeatures(int numSelFeatures, ParallelTopicModel lda, Alphabet alphabet) {
ArrayList<Integer> features = new ArrayList<Integer>();
Alphabet seqAlphabet = lda.getAlphabet();
int numTopics = lda.getNumTopics();
Object[][] sorted = lda.getTopWords(seqAlphabet.size());
for (int pos = 0; pos < seqAlphabet.size(); pos++) {
for (int ti = 0; ti < numTopics; ti++) {
Object feat = sorted[ti][pos].toString();
int fi = alphabet.lookupIndex(feat,false);
if ((fi >=0) && (!features.contains(fi))) {
logger.info("Selected feature: " + feat);
features.add(fi);
if (features.size() == numSelFeatures) {
return features;
}
}
}
}
return features;
}