Full fix for #64

src/NlpTools/Models/Lda.php

@@ -225,7 +225,7 @@ public function getPhi($limit_words=-1)
       * Get the probability of a document given a topic (theta according
       * to Griffiths and Steyvers)
       *
-      * @param $limit_docs Limit the results to the top n docs
+      * @param $limit_docs Limit the results to the top n words
       * @return array A two dimensional array that contains the probabilities for each document
       */
      public function getDocumentsPerTopicsProbabilities($limit_docs=-1)
@@ -239,21 +239,28 @@ public function getDocumentsPerTopicsProbabilities($limit_docs=-1)
          $denom = $doccnt + $this->ntopics*$this->a;
          $count_topics_docs = array();
          foreach ($this->count_docs_topics as $doc=>$topics) {
-             foreach ($topics as $t=>$c)
-                $count_topics_docs[$doc][$t]++;
+             foreach ($topics as $t=>$c) {
+                 if (!isset($count_topics_docs[$doc])) {
+                     $count_topics_docs[$doc] = array();
+                 }
+                 if (!isset($count_topics_docs[$doc][$t])) {
+                     $count_topics_docs[$doc][$t] = 0;
+                 }
+                 $count_topics_docs[$doc][$t]+=$c;
+             }
          }
 
          foreach ($p_t_d as $topic=>&$p) {
              foreach ($count_topics_docs as $doc=>$tc) {
                  $p[$doc] = ($tc[$topic] + $this->a)/$denom;
              }
-             if ($limit_words>0) {
+             if ($limit_docs>0) {
                  arsort($p);
-                 $p = array_slice($p,0,$limit_words,true); // true to preserve the keys
+                 $p = array_slice($p,0,$limit_docs,true); // true to preserve the keys
              }
          }
 
-         return $p;
+         return $p_t_d;
      }
 
      /**