added an option to use a uniform prior rather than a learnt one

lib/classificator.js

@@ -20,6 +20,7 @@ const STATE_KEYS = (module.exports.STATE_KEYS = [
   'options',
 ]);
 const DEFAULT_ALPHA = 1
+const DEFAULT_FIT_PRIOR = true
 
 /**
  * Initializes a NaiveBayes instance from a JSON state representation.
@@ -107,7 +108,7 @@ function Naivebayes(options) {
 
   this.tokenizer = this.options.tokenizer || defaultTokenizer;
   this.alpha = this.options.alpha || DEFAULT_ALPHA
-
+  this.fitPrior = this.options.fitPrior === undefined ? DEFAULT_FIT_PRIOR : this.options.fitPrior
   //initialize our vocabulary and its size
   this.vocabulary = {};
   this.vocabularySize = 0;
@@ -295,7 +296,12 @@ Naivebayes.prototype.categorize = function(text){
     //start by calculating the overall probability of this category
     //=>  out of all documents we've ever looked at, how many were
     //    mapped to this category
-    let categoryLikelihood = this.docCount[category] / this.totalDocuments;
+    let categoryLikelihood
+    if (this.fitPrior) {
+      categoryLikelihood = this.docCount[category] / this.totalDocuments;
+    } else {
+      categoryLikelihood = 1
+    }
 
     //take the log to avoid underflow
     // let logLikelihood = Math.log(categoryLikelihood);