feat: Allows a Pastel model to be created from a dict of feature names to weights

src/pastel/pastel.py

@@ -95,15 +95,11 @@ def from_feature_list(feature_names: Sequence[FEATURE_TYPE]) -> "Pastel":
         return Pastel(new_model)
 
     @staticmethod
-    def load_model(model_file: str) -> "Pastel":
-        """Load model from JSON file. Convert any functions in the model
-        from their names to Callable functions."""
-
-        with open(model_file, "rt", encoding="utf-8") as json_in:
-            model_json = json.load(json_in)
+    def from_dict(model_dict: dict[str, float]) -> "Pastel":
+        "Create model from a map of features to weights"
         # replace function names with function objects found in pastel_functions module
         new_model = {}
-        for feature, weight in model_json.items():
+        for feature, weight in model_dict.items():
             if feature in pastel_functions.__all__:
                 new_model[getattr(pastel_functions, feature)] = weight
             elif feature == "bias":
@@ -113,6 +109,15 @@ def load_model(model_file: str) -> "Pastel":
 
         return Pastel(new_model)
 
+    @staticmethod
+    def load_model(model_file: str) -> "Pastel":
+        """Load model from JSON file. Convert any functions in the model
+        from their names to Callable functions."""
+
+        with open(model_file, "rt", encoding="utf-8") as json_in:
+            model_json = json.load(json_in)
+        return Pastel.from_dict(model_json)
+
     def save_model(self, model_path: str) -> None:
         """
         Save the questions, functions and associated weights to a local JSON file
@@ -250,7 +255,7 @@ async def get_answers_to_questions(
     ) -> dict[Sentence, dict[FEATURE_TYPE, float]]:
         """Embed each example into the prompt and pass to genAI, then
         get answers for non-genAI functions.
-        For each sentence, this Returns a dictionary mapping features to scores."""
+        For each sentence, this returns a dictionary mapping features to scores."""
 
         jobs = [
             self._get_answers_for_single_sentence(sentence) for sentence in sentences

tests/pastel/test_pastel.py

@@ -36,6 +36,14 @@ def test_load_file(pastel_instance: Pastel) -> None:
     assert loaded.model == pastel_instance.model
 
 
+def test_model_from_dict():
+    model_dict = {"bias": 1.0, "question_0?": 2.0, "is_claim_type_personal": 3.0}
+    model = Pastel.from_dict(model_dict)
+    assert model.get_bias() == 1.0
+    assert len(model.get_functions()) == 1
+    assert len(model.get_questions()) == 1
+
+
 def test_make_prompt(pastel_instance: Pastel) -> None:
     sentence = Sentence("The sky is blue.", tuple("quantity"))
     prompt = pastel_instance.make_prompt(sentence)