Merge pull request #1 from ngirase10/feat/categorical-update-categories

Add optional parameter for factorize to pass in original codes, uniques

pandas/core/algorithms.py

@@ -633,6 +633,7 @@ def factorize(
     sort: bool = False,
     use_na_sentinel: bool = True,
     size_hint: int | None = None,
+    original_factorization: tuple[np.ndarray, np.ndarray | Index] | None = None
 ) -> tuple[np.ndarray, np.ndarray | Index]:
     """
     Encode the object as an enumerated type or categorical variable.
@@ -758,6 +759,28 @@ def factorize(
     # Step 2 is dispatched to extension types (like Categorical). They are
     # responsible only for factorization. All data coercion, sorting and boxing
     # should happen here.
+
+    if original_factorization is not None:
+        original_uniques, original_codes = original_factorization
+        unique_to_code = dict(zip(original_uniques, range(len(original_uniques))))
+
+        # Map existing data to original codes, assign new codes to new uniques
+        new_codes = []
+        new_uniques = list(original_uniques)
+        next_code = len(original_uniques)
+
+        for item in values:
+            if item in unique_to_code:
+                new_codes.append(unique_to_code[item])
+            else:
+                unique_to_code[item] = next_code
+                new_uniques.append(item)
+                new_codes.append(next_code)
+                next_code += 1
+
+        return new_codes, new_uniques
+
+
     if isinstance(values, (ABCIndex, ABCSeries)):
         return values.factorize(sort=sort, use_na_sentinel=use_na_sentinel)
 

pandas/core/arrays/categorical.py

@@ -679,6 +679,28 @@ def _from_inferred_categories(
 
         return cls._simple_new(codes, dtype=dtype)
 
+    @classmethod
+    def update_categories(self, new_categories):
+
+        print("Debug - self.categories:", self.categories)
+
+
+        # Ensure new_categories is a list
+        if not isinstance(new_categories, list):
+            raise ValueError("new_categories must be a list")
+
+        # Combine existing and new categories, maintaining order and uniqueness
+        current_categories = self.categories
+        print("self.categories", self.categories)
+        print("current", current_categories)
+        updated_categories = list(current_categories)
+        for cat in new_categories:
+            if cat not in updated_categories:
+                updated_categories.append(cat)
+
+        # Update the categories in-place
+        self.set_categories(updated_categories, inplace=True)
+
     @classmethod
     def from_codes(
         cls,