本文整理汇总了Java中edu.jhu.prim.util.math.FastMath.mod方法的典型用法代码示例。如果您正苦于以下问题：Java FastMath.mod方法的具体用法？Java FastMath.mod怎么用？Java FastMath.mod使用的例子？那么恭喜您, 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edu.jhu.prim.util.math.FastMath的用法示例。
在下文中一共展示了FastMath.mod方法的5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: addFeat
import edu.jhu.prim.util.math.FastMath; //导入方法依赖的package包/类
private static void addFeat(FeatureVector feats, int mod, long feat) {
int hash = MurmurHash.hash32(feat);
if (mod > 0) {
hash = FastMath.mod(hash, mod);
}
feats.add(hash, 1.0);
// Enable this for debugging of feature creation.
// if (feats instanceof LongFeatureVector) {
// ((LongFeatureVector)feats).addLong(feat, 1.0);
// }
}
示例2: getValIdx
import edu.jhu.prim.util.math.FastMath; //导入方法依赖的package包/类
public int getValIdx(AnnoSentence sent, int pidx, int cidx) {
String val = getVal(sent, pidx, cidx);
if (val == null) {
return -1;
} else {
String data = val;
if (valueHashMod > 0) {
int idx = FastMath.mod(MurmurHash3.murmurhash3_x86_32(data, 0, data.length(), 123456789), valueHashMod);
return valAlphabet.lookupIndex(idx);
} else {
return valAlphabet.lookupIndex(val);
}
}
}
示例3: addFeat
import edu.jhu.prim.util.math.FastMath; //导入方法依赖的package包/类
public static void addFeat(FeatureVector feats, int mod, long feat, double value) {
int hash = MurmurHash.hash32(feat);
if (mod > 0) {
hash = FastMath.mod(hash, mod);
}
feats.add(hash, value);
}
示例4: getFeatures
import edu.jhu.prim.util.math.FastMath; //导入方法依赖的package包/类
protected static void getFeatures(List<FeatTemplate> tpls, TemplateFeatureExtractor extStr, IntTemplateFeatureExtractor extInt,
LocalObservations local) {
for (FeatTemplate tpl : tpls) {
List<String> featsStr = new ArrayList<>();
extStr.addFeatures(tpl, local, featsStr);
IntArrayList featsInt = new IntArrayList();
extInt.addFeatures(tpl, local, featsInt);
if (featsStr.size() != featsInt.size()) {
log.error("Mismatch in number of features extracted for template: {} str = {} int = {}", tpl.getName(), featsStr.size(), featsInt.size());
}
for (int i=0; i<Math.min(featsStr.size(), featsInt.size()); i++) {
int hash;
if (useIntExtr) {
hash = featsInt.get(i);
} else {
hash = MurmurHash.hash32(featsStr.get(i));
}
hash = FastMath.mod(hash, featureHashMod);
// Add to the set of strings which are colliding on this hash.
Set<String> values = collisions[hash];
if (values == null) {
values = new HashSet<String>();
collisions[hash] = values;
}
values.add(featsStr.get(i));
// Increment the token count of collisions on this hash.
collisionTokenCount[hash] += 1;
}
}
}
示例5: computeInformationGain
import edu.jhu.prim.util.math.FastMath; //导入方法依赖的package包/类
private void computeInformationGain(int t, List<FeatTemplate> allTpls, List<ValExtractor> valExts,
AnnoSentenceCollection inputSents, AnnoSentenceCollection goldSents, CorpusStatistics cs, double[][] ig, int[] featCount) {
FeatTemplate tpl = allTpls.get(t);
final IntDoubleDenseVector[][] counts = getCountsArray(valExts);
IntObjectBimap<String> alphabet = new IntObjectBimap<String>();
for (int i=0; i<goldSents.size(); i++) {
AnnoSentence goldSent = goldSents.get(i);
AnnoSentence inputSent = inputSents.get(i);
TemplateFeatureExtractor featExt = new TemplateFeatureExtractor(inputSent, cs);
for (int pidx=-1; pidx<inputSent.size(); pidx++) {
for (int cidx=0; cidx<inputSent.size(); cidx++) {
// Feature Extraction.
List<String> feats = new ArrayList<String>();
featExt.addFeatures(QLists.getList(tpl), LocalObservations.newPidxCidx(pidx, cidx), feats);
if (feats.size() == 0) {
if (!tplsWithNoFeats.contains(tpl)) {
log.warn("No features extracted for template: " + tpl.getName());
tplsWithNoFeats.add(tpl);
}
}
FeatureVector fv = new FeatureVector();
for (int j=0; j<feats.size(); j++) {
String featName = feats.get(j);
int featIdx;
if (prm.featureHashMod > 0) {
String data = featName;
featIdx = FastMath.mod(MurmurHash3.murmurhash3_x86_32(data, 0, data.length(), 123456789), prm.featureHashMod);
featIdx = alphabet.lookupIndex(Integer.toString(featIdx));
} else {
featIdx = alphabet.lookupIndex(featName);
}
fv.add(featIdx, 1.0);
}
// For each value extractor:
for (int c = 0; c < valExts.size(); c++) {
// Value Extraction.
ValExtractor valExt = valExts.get(c);
final int valIdx = valExt.getValIdx(goldSent, pidx, cidx);
if (valIdx != -1) {
// Increment counts of feature and value occurrences.
counts[c][valIdx].add(fv);
}
}
}
}
}
for (int c = 0; c < valExts.size(); c++) {
// Compute information gain for this (feature template, value extractor) pair.
ig[c][t] = computeInformationGain(counts[c]);
}
featCount[t] = alphabet.size();
if (t % 10 == 0) {
log.debug(String.format("Processed feature template %d of %d: %s #feats=%d", t, allTpls.size(), tpl.getName(), alphabet.size()));
}
}