@sc-voice/tools 2.11.0 → 2.13.0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@sc-voice/tools",
-  "version": "2.11.0",
+  "version": "2.13.0",
   "description": "Utilities for SC-Voice",
   "main": "index.mjs",
   "files": [

package/src/text/tfidf-space.mjs

@@ -70,41 +70,37 @@ export class TfidfSpace {
     return TfidfSpace.removeNonWords(sAbbr);
   }
 
-  static idfStandard(space, word) {
-    const msg = 'w7e.idfStandard:';
-    let { corpus } = space;
-    let wordDocs = corpus.wordDocCount[word] || 0;
-    return Math.log((corpus.size + 1) / (wordDocs + 1));
+  static idfStandard(nDocs, wdc, idfWeight) {
+    return Math.log((nDocs + 1) / (wdc + 1));
   }
 
-  static idfTunable(space, word, idfWeight) {
-    const msg = 'w7e.idf:';
-    let { corpus } = space;
-    let wordDocs = corpus.wordDocCount[word] || 0;
+  static idfTunable(nDocs, wdc, idfWeight) {
+    const msg = 'w7e.idfTunable:';
     // NOTE: This is NOT the usual formula
     // Map to [0:ignore..1:important]
-    return corpus.size
-      ? 1 -
-          Math.exp(((wordDocs - corpus.size) / wordDocs) * idfWeight)
+    return nDocs
+      ? 1 - Math.exp(((wdc - nDocs) / wdc) * idfWeight)
       : 1;
   }
 
   idf(word, idfWeight = this.idfWeight) {
-    return this.idfFunction(this, word, idfWeight);
+    let { corpus } = this;
+    let wdc = corpus.wordDocCount[word] || 0;
+    let nDocs = corpus.size;
+    return this.idfFunction(nDocs, wdc, idfWeight);
   }
 
-  addCorpusDocument(id, bow, nWords) {
+  addCorpusDocument(id, bow) {
     const msg = 't8w.addCorpusDocument:';
     let { corpus } = this;
     if (id == null) {
       throw new Error(`${msg} id?`);
     }
     if (bow == null) {
+      // Bag-of-words maps word to wordCount(word,doc)
       throw new Error(`${msg} bow?`);
     }
-    if (nWords == null) {
-      throw new Error(`${msg} nWords?`);
-    }
+    let nWords = Object.values(bow).reduce((a,v)=>a+v);
     let docInfo = { id, bow, nWords };
     corpus.wordDocCount.increment(bow.oneHot());
     corpus.addDocument(id, docInfo);
@@ -129,20 +125,20 @@ export class TfidfSpace {
     return count ? count / words.length : 0;
   }
 
-  tfidf(doc) {
-    const msg = 'w7e.tfidf:';
+  tfidfOfBow(bow) {
+    const msg = 'w7e.tfidfOfBow:';
     let { corpus, idfWeight } = this;
 
     // More efficient implementation of tf * idf
-    let { bow, words } = this.countWords(doc);
-    let nWords = words.length;
+    let words = Object.keys(bow);
+    let nWords = words.reduce((a,w)=>a+bow[w],0);
 
     let vTfIdf = words.reduce((a, word) => {
       let wd = bow[word] || 0;
       let tf = wd ? wd / nWords : 0;
-      let wc = corpus.wordDocCount[word] || 0;
+      let wdc = corpus.wordDocCount[word] || 0;
       let idf = corpus.size
-        ? 1 - Math.exp(((wc - corpus.size) / wc) * idfWeight)
+        ? 1 - Math.exp(((wdc - corpus.size) / wdc) * idfWeight)
         : 1;
       let tfidf = tf * idf;
       if (tfidf) {
@@ -154,6 +150,11 @@ export class TfidfSpace {
     return vTfIdf;
   }
 
+  tfidf(text) { // TfIdf of words in text w/r to corpus
+    let { bow } = this.countWords(text);
+    return this.tfidfOfBow(bow);
+  }
+
   countWords(str) {
     const msg = 'w7e.countWords:';
     if (str == null) {