feat: renamed sample operations and cleaned up test warnings (#1407)

this is a continuation of #1394 
cleaned up all the pydantic issues because of this

Author: jjmachan

pyproject.toml

@@ -59,4 +59,5 @@ build-backend = "setuptools.build_meta"
 write_to = "src/ragas/_version.py"
 
 [tool.pytest.ini_options]
-addopts = "-n 4"
+addopts = "-n 4"
+asyncio_default_fixture_loop_scope = "function"

src/ragas/dataset_schema.py

@@ -13,16 +13,24 @@
 
 
 class BaseEvalSample(BaseModel):
-    def dict(self, **kwargs):
-        row = super().dict(**kwargs)
-        row = {k: v for k, v in row.items() if v is not None}
-        return row
+    def to_dict(self) -> t.Dict:
+        """
+        Get the dictionary representation of the sample without attributes that are None.
+        """
+        return self.model_dump(exclude_none=True)
 
-    def features(self):
-        return set(self.dict().keys())
+    def get_features(self) -> t.List[str]:
+        """
+        Get the features of the sample that are not None.
+        """
+        return list(self.to_dict().keys())
 
 
 class SingleTurnSample(BaseEvalSample):
+    """
+    Represents evaluation samples for single-turn interactions.
+    """
+
     user_input: t.Optional[str] = None
     retrieved_contexts: t.Optional[t.List[str]] = None
     reference_contexts: t.Optional[t.List[str]] = None
@@ -68,7 +76,7 @@ def validate_user_input(
         return messages
 
     def to_messages(self):
-        return [m.dict() for m in self.user_input]
+        return [m.model_dump() for m in self.user_input]
 
     def pretty_repr(self):
         lines = []
@@ -98,7 +106,7 @@ def get_sample_type(self):
         return type(self.samples[0])
 
     def _to_list(self) -> t.List[t.Dict]:
-        rows = [sample.dict() for sample in self.samples]
+        rows = [sample.model_dump() for sample in self.samples]
 
         if self.get_sample_type() == MultiTurnSample:
             for sample in rows:
@@ -130,7 +138,7 @@ def to_pandas(self) -> PandasDataframe:
         return pd.DataFrame(data)
 
     def features(self):
-        return self.samples[0].features()
+        return self.samples[0].get_features()
 
     @classmethod
     def from_list(cls, mapping: t.List[t.Dict]):

src/ragas/experimental/metrics/_faithfulness.py

@@ -248,5 +248,5 @@ async def _ascore(self, row: t.Dict, callbacks: Callbacks) -> float:
     async def _single_turn_ascore(
         self: t.Self, sample: SingleTurnSample, callbacks: Callbacks
     ) -> float:
-        row = sample.dict()
+        row = sample.to_dict()
         return await self._ascore(row, callbacks)

src/ragas/integrations/langchain.py

@@ -48,9 +48,9 @@ def __init__(self, metric: Metric, **kwargs: t.Any):
             t.cast(MetricWithLLM, self.metric).llm = LangchainLLMWrapper(llm)
         if isinstance(self.metric, MetricWithEmbeddings):
             embeddings = get_or_init(kwargs, "embeddings", OpenAIEmbeddings)
-            t.cast(MetricWithEmbeddings, self.metric).embeddings = (
-                LangchainEmbeddingsWrapper(embeddings)
-            )
+            t.cast(
+                MetricWithEmbeddings, self.metric
+            ).embeddings = LangchainEmbeddingsWrapper(embeddings)
         self.metric.init(run_config)
 
         assert isinstance(
@@ -132,7 +132,7 @@ def _validate(self, input: SingleTurnSample) -> None:
         # validate each example
         required_columns = self.metric.required_columns.get("SINGLE_TURN", [])
         for col in required_columns:
-            if col not in input.features():
+            if col not in input.get_features():
                 raise ValueError(
                     f'"{col}" is required in each example'
                     f"for the metric[{self.metric.name}] you have chosen."

src/ragas/metrics/_answer_correctness.py

@@ -219,7 +219,7 @@ def _create_statements_prompt(self, question: str, text: str) -> PromptValue:
     async def _single_turn_ascore(
         self: t.Self, sample: SingleTurnSample, callbacks: Callbacks
     ) -> float:
-        row = sample.dict()
+        row = sample.to_dict()
         score = await self._ascore(row, callbacks)
         return score
 

src/ragas/metrics/_answer_relevance.py

@@ -160,7 +160,7 @@ def _create_question_gen_prompt(self, row: t.Dict) -> PromptValue:
     async def _single_turn_ascore(
         self, sample: SingleTurnSample, callbacks: Callbacks
     ) -> float:
-        row = sample.dict()
+        row = sample.to_dict()
         return await self._ascore(row, callbacks)
 
     async def _ascore(self, row: t.Dict, callbacks: Callbacks) -> float:

src/ragas/metrics/_answer_similarity.py

@@ -60,7 +60,7 @@ def __post_init__(self: t.Self):
     async def _single_turn_ascore(
         self, sample: SingleTurnSample, callbacks: Callbacks
     ) -> float:
-        row = sample.dict()
+        row = sample.to_dict()
         return await self._ascore(row, callbacks)
 
     async def _ascore(self: t.Self, row: t.Dict, callbacks: Callbacks) -> float:

src/ragas/metrics/_aspect_critic.py

@@ -143,7 +143,7 @@ def _compute_score(
     async def _single_turn_ascore(
         self: t.Self, sample: SingleTurnSample, callbacks: Callbacks
     ) -> float:
-        row = sample.dict()
+        row = sample.to_dict()
         return await self._ascore(row, callbacks)
 
     async def _ascore(self: t.Self, row: t.Dict, callbacks: Callbacks) -> float:

src/ragas/metrics/_context_entities_recall.py

@@ -175,7 +175,7 @@ async def get_entities(
     async def _single_turn_ascore(
         self, sample: SingleTurnSample, callbacks: Callbacks
     ) -> float:
-        row = sample.dict()
+        row = sample.to_dict()
         return await self._ascore(row, callbacks)
 
     async def _ascore(

src/ragas/metrics/_context_precision.py

@@ -154,7 +154,7 @@ def _calculate_average_precision(
     async def _single_turn_ascore(
         self, sample: SingleTurnSample, callbacks: Callbacks
     ) -> float:
-        row = sample.dict()
+        row = sample.to_dict()
         return await self._ascore(row, callbacks)
 
     async def _ascore(