Compiler performance improvements

gptc/compiler.py

@@ -41,39 +41,30 @@ def compile(
 
     word_counts: Dict[int, Dict[str, int]] = {}
 
-    names = []
+    names = tuple(categories.keys())
 
     for category, text in categories.items():
-        if not category in names:
-            names.append(category)
-
         for word in text:
-            try:
+            if word in word_counts:
-                counts_for_word = word_counts[word]
+                try:
-            except KeyError:
+                    word_counts[word][category] += 1
-                counts_for_word = {}
+                except KeyError:
-                word_counts[word] = counts_for_word
+                    word_counts[word][category] = 1
+            else:
+                word_counts[word] = {category: 1}
 
-            try:
+    category_lengths = {
-                word_counts[word][category] += 1
+        category: len(text) for category, text in categories.items()
-            except KeyError:
-                word_counts[word][category] = 1
-
-    word_counts = {
-        word: counts
-        for word, counts in word_counts.items()
-        if sum(counts.values()) >= min_count
     }
 
-    word_weights: Dict[int, Dict[str, float]] = {}
+    word_weights: Dict[int, Dict[str, float]] = {
-    for word, values in word_counts.items():
+        word: {
-        for category, value in values.items():
+            category: value / category_lengths[category]
-            try:
+            for category, value in values.items()
-                word_weights[word][category] = value / len(categories[category])
+        }
-            except KeyError:
+        for word, values in word_counts.items()
-                word_weights[word] = {
+        if sum(values.values()) >= min_count
-                    category: value / len(categories[category])
+    }
-                }
 
     model: Dict[int, List[int]] = {}
     for word, weights in word_weights.items():