模式

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 get_features(self) -> t.List[str]:
    """
    Get the features of the sample that are not None.
    """
    return list(self.to_dict().keys())

def to_string(self) -> str:
    """
    Get the string representation of the sample.
    """
    sample_dict = self.to_dict()
    return "".join(f"\n{key}:\n\t{val}\n" for key, val in sample_dict.items())

@field_validator("user_input")
@classmethod
def validate_user_input(
    cls,
    messages: t.List[t.Union[HumanMessage, AIMessage, ToolMessage]],
) -> t.List[t.Union[HumanMessage, AIMessage, ToolMessage]]:
    """Validates the user input messages."""
    if not (
        isinstance(m, (HumanMessage, AIMessage, ToolMessage)) for m in messages
    ):
        raise ValueError(
            "All inputs must be instances of HumanMessage, AIMessage, or ToolMessage."
        )

    has_seen_ai_message = False

    for i, m in enumerate(messages):
        if isinstance(m, AIMessage):
            has_seen_ai_message = True

        elif isinstance(m, ToolMessage):
            # Rule 1: ToolMessage must be preceded by an AIMessage somewhere in the conversation
            if not has_seen_ai_message:
                raise ValueError(
                    "ToolMessage must be preceded by an AIMessage somewhere in the conversation."
                )

            # Rule 2: ToolMessage must follow an AIMessage or another ToolMessage
            if i > 0:
                prev_message = messages[i - 1]

                if isinstance(prev_message, AIMessage):
                    # Rule 3: If following AIMessage, that message must have tool_calls
                    if not prev_message.tool_calls:
                        raise ValueError(
                            "ToolMessage must follow an AIMessage where tools were called."
                        )
                elif not isinstance(prev_message, ToolMessage):
                    # Not following AIMessage or ToolMessage
                    raise ValueError(
                        "ToolMessage must follow an AIMessage or another ToolMessage."
                    )

    return messages

def to_messages(self):
    """Converts the user input messages to a list of dictionaries."""
    return [m.model_dump() for m in self.user_input]

def pretty_repr(self):
    """Returns a pretty string representation of the conversation."""
    lines = []
    for m in self.user_input:
        lines.append(m.pretty_repr())

    return "\n".join(lines)

@abstractmethod
def to_list(self) -> t.List[t.Dict]:
    """Converts the samples to a list of dictionaries."""
    pass

@classmethod
@abstractmethod
def from_list(cls: t.Type[T], data: t.List[t.Dict]) -> T:
    """Creates an RagasDataset from a list of dictionaries."""
    pass

def validate_samples(self, samples: t.List[Sample]) -> t.List[Sample]:
    """Validates that all samples are of the same type."""
    if len(samples) == 0:
        return samples

    first_sample_type = type(samples[0])
    for i, sample in enumerate(samples):
        if not isinstance(sample, first_sample_type):
            raise ValueError(
                f"Sample at index {i} is of type {type(sample)}, expected {first_sample_type}"
            )

    return samples

def get_sample_type(self) -> t.Type[Sample]:
    """Returns the type of the samples in the dataset."""
    return type(self.samples[0])

def to_hf_dataset(self) -> HFDataset:
    """Converts the dataset to a Hugging Face Dataset."""
    try:
        from datasets import Dataset as HFDataset
    except ImportError:
        raise ImportError(
            "datasets is not installed. Please install it to use this function."
        )

    return HFDataset.from_list(self.to_list())

@classmethod
def from_hf_dataset(cls: t.Type[T], dataset: HFDataset) -> T:
    """Creates an EvaluationDataset from a Hugging Face Dataset."""
    return cls.from_list(dataset.to_list())

def to_pandas(self) -> PandasDataframe:
    """Converts the dataset to a pandas DataFrame."""
    try:
        import pandas as pd
    except ImportError:
        raise ImportError(
            "pandas is not installed. Please install it to use this function."
        )

    data = self.to_list()
    return pd.DataFrame(data)

@classmethod
def from_pandas(cls, dataframe: PandasDataframe):
    """Creates an EvaluationDataset from a pandas DataFrame."""
    return cls.from_list(dataframe.to_dict(orient="records"))

def features(self):
    """Returns the features of the samples."""
    return self.samples[0].get_features()

@classmethod
def from_dict(cls: t.Type[T], mapping: t.Dict) -> T:
    """Creates an EvaluationDataset from a dictionary."""
    samples = []
    if all(
        "user_input" in item and isinstance(mapping[0]["user_input"], list)
        for item in mapping
    ):
        samples.extend(MultiTurnSample(**sample) for sample in mapping)
    else:
        samples.extend(SingleTurnSample(**sample) for sample in mapping)
    return cls(samples=samples)

def to_csv(self, path: t.Union[str, Path]):
    """Converts the dataset to a CSV file."""
    import csv

    data = self.to_list()
    if not data:
        return

    fieldnames = data[0].keys()

    with open(path, "w", newline="") as csvfile:
        writer = csv.DictWriter(csvfile, fieldnames=fieldnames)
        writer.writeheader()
        for row in data:
            writer.writerow(row)

def to_jsonl(self, path: t.Union[str, Path]):
    """Converts the dataset to a JSONL file."""
    with open(path, "w") as jsonlfile:
        for sample in self.to_list():
            jsonlfile.write(json.dumps(sample, ensure_ascii=False) + "\n")

@classmethod
def from_jsonl(cls: t.Type[T], path: t.Union[str, Path]) -> T:
    """Creates an EvaluationDataset from a JSONL file."""
    with open(path, "r") as jsonlfile:
        data = [json.loads(line) for line in jsonlfile]
    return cls.from_list(data)

def to_pandas(self, batch_size: int | None = None, batched: bool = False):
    """
    Convert the result to a pandas DataFrame.

    Parameters
    ----------
    batch_size : int, optional
        The batch size for conversion. Default is None.
    batched : bool, optional
        Whether to convert in batches. Default is False.

    Returns
    -------
    pandas.DataFrame
        The result as a pandas DataFrame.

    Raises
    ------
    ValueError
        If the dataset is not provided.
    """
    try:
        import pandas as pd
    except ImportError:
        raise ImportError(
            "pandas is not installed. Please install it to use this function."
        )

    if self.dataset is None:
        raise ValueError("dataset is not provided for the results class")
    assert len(self.scores) == len(self.dataset)
    # convert both to pandas dataframes and concatenate
    scores_df = pd.DataFrame(self.scores)
    dataset_df = self.dataset.to_pandas()
    return pd.concat([dataset_df, scores_df], axis=1)

def total_tokens(self) -> t.Union[t.List[TokenUsage], TokenUsage]:
    """
    Compute the total tokens used in the evaluation.

    Returns
    -------
    list of TokenUsage or TokenUsage
        The total tokens used.

    Raises
    ------
    ValueError
        If the cost callback handler is not provided.
    """
    if self.cost_cb is None:
        raise ValueError(
            "The evaluate() run was not configured for computing cost. Please provide a token_usage_parser function to evaluate() to compute cost."
        )
    return self.cost_cb.total_tokens()

def total_cost(
    self,
    cost_per_input_token: t.Optional[float] = None,
    cost_per_output_token: t.Optional[float] = None,
    per_model_costs: t.Dict[str, t.Tuple[float, float]] = {},
) -> float:
    """
    Compute the total cost of the evaluation.

    Parameters
    ----------
    cost_per_input_token : float, optional
        The cost per input token. Default is None.
    cost_per_output_token : float, optional
        The cost per output token. Default is None.
    per_model_costs : dict of str to tuple of float, optional
        The per model costs. Default is an empty dictionary.

    Returns
    -------
    float
        The total cost of the evaluation.

    Raises
    ------
    ValueError
        If the cost callback handler is not provided.
    """
    if self.cost_cb is None:
        raise ValueError(
            "The evaluate() run was not configured for computing cost. Please provide a token_usage_parser function to evaluate() to compute cost."
        )
    return self.cost_cb.total_cost(
        cost_per_input_token, cost_per_output_token, per_model_costs
    )

@classmethod
def from_json(cls, path: str, metric_name: t.Optional[str]) -> "MetricAnnotation":
    """Load annotations from a JSON file"""
    dataset = json.load(open(path))
    return cls._process_dataset(dataset, metric_name)

@classmethod
def from_app(
    cls,
    run_id: str,
    metric_name: t.Optional[str] = None,
) -> "MetricAnnotation":
    """
    Fetch annotations from a URL using either evaluation result or run_id

    Parameters
    ----------
    run_id : str
        Direct run ID to fetch annotations
    metric_name : str, optional
        Name of the specific metric to filter

    Returns
    -------
    MetricAnnotation
        Annotation data from the API

    Raises
    ------
    ValueError
        If run_id is not provided
    """
    if run_id is None:
        raise ValueError("run_id must be provided")

    endpoint = f"/api/v1/alignment/evaluation/annotation/{run_id}"

    app_token = get_app_token()
    base_url = get_api_url()
    app_url = get_app_url()

    response = requests.get(
        f"{base_url}{endpoint}",
        headers={
            "Content-Type": "application/json",
            "x-app-token": app_token,
            "x-source": RAGAS_API_SOURCE,
            "x-app-version": __version__,
        },
    )

    check_api_response(response)
    dataset = response.json()["data"]

    if not dataset:
        evaluation_url = build_evaluation_app_url(app_url, run_id)
        raise ValueError(
            f"No annotations found. Please annotate the Evaluation first then run this method. "
            f"\nNote: you can annotate the evaluations using the Ragas app by going to {evaluation_url}"
        )

    return cls._process_dataset(dataset, metric_name)

def train_test_split(
    self,
    test_size: float = 0.2,
    seed: int = 42,
    stratify: t.Optional[t.List[t.Any]] = None,
) -> t.Tuple["SingleMetricAnnotation", "SingleMetricAnnotation"]:
    """
    Split the dataset into training and testing sets.

    Parameters:
        test_size (float): The proportion of the dataset to include in the test split.
        seed (int): Random seed for reproducibility.
        stratify (list): The column values to stratify the split on.
    """
    raise NotImplementedError

def sample(
    self, n: int, stratify_key: t.Optional[str] = None
) -> "SingleMetricAnnotation":
    """
    Create a subset of the dataset.

    Parameters:
        n (int): The number of samples to include in the subset.
        stratify_key (str): The column to stratify the subset on.

    Returns:
        SingleMetricAnnotation: A subset of the dataset with `n` samples.
    """
    if n > len(self.samples):
        raise ValueError(
            "Requested sample size exceeds the number of available samples."
        )

    if stratify_key is None:
        # Simple random sampling
        sampled_indices = random.sample(range(len(self.samples)), n)
        sampled_samples = [self.samples[i] for i in sampled_indices]
    else:
        # Stratified sampling
        class_groups = defaultdict(list)
        for idx, sample in enumerate(self.samples):
            key = sample[stratify_key]
            class_groups[key].append(idx)

        # Determine the proportion of samples to take from each class
        total_samples = sum(len(indices) for indices in class_groups.values())
        proportions = {
            cls: len(indices) / total_samples
            for cls, indices in class_groups.items()
        }

        sampled_indices = []
        for cls, indices in class_groups.items():
            cls_sample_count = int(np.round(proportions[cls] * n))
            cls_sample_count = min(
                cls_sample_count, len(indices)
            )  # Don't oversample
            sampled_indices.extend(random.sample(indices, cls_sample_count))

        # Handle any rounding discrepancies to ensure exactly `n` samples
        while len(sampled_indices) < n:
            remaining_indices = set(range(len(self.samples))) - set(sampled_indices)
            if not remaining_indices:
                break
            sampled_indices.append(random.choice(list(remaining_indices)))

        sampled_samples = [self.samples[i] for i in sampled_indices]

    return SingleMetricAnnotation(name=self.name, samples=sampled_samples)

def batch(
    self,
    batch_size: int,
    drop_last_batch: bool = False,
):
    """
    Create a batch iterator.

    Parameters:
        batch_size (int): The number of samples in each batch.
        stratify (str): The column to stratify the batches on.
        drop_last_batch (bool): Whether to drop the last batch if it is smaller than the specified batch size.
    """

    samples = self.samples[:]
    random.shuffle(samples)

    all_batches = [
        samples[i : i + batch_size]
        for i in range(0, len(samples), batch_size)
        if len(samples[i : i + batch_size]) == batch_size or not drop_last_batch
    ]

    return all_batches

def stratified_batches(
    self,
    batch_size: int,
    stratify_key: str,
    drop_last_batch: bool = False,
    replace: bool = False,
) -> t.List[t.List[SampleAnnotation]]:
    """
    Create stratified batches based on a specified key, ensuring proportional representation.

    Parameters:
        batch_size (int): Number of samples per batch.
        stratify_key (str): Key in `metric_input` used for stratification (e.g., class labels).
        drop_last_batch (bool): If True, drops the last batch if it has fewer samples than `batch_size`.
        replace (bool): If True, allows reusing samples from the same class to fill a batch if necessary.

    Returns:
        List[List[SampleAnnotation]]: A list of stratified batches, each batch being a list of SampleAnnotation objects.
    """
    # Group samples based on the stratification key
    class_groups = defaultdict(list)
    for sample in self.samples:
        key = sample[stratify_key]
        class_groups[key].append(sample)

    # Shuffle each class group for randomness
    for group in class_groups.values():
        random.shuffle(group)

    # Determine the number of batches required
    total_samples = len(self.samples)
    num_batches = (
        np.ceil(total_samples / batch_size).astype(int)
        if drop_last_batch
        else np.floor(total_samples / batch_size).astype(int)
    )
    samples_per_class_per_batch = {
        cls: max(1, len(samples) // num_batches)
        for cls, samples in class_groups.items()
    }

    # Create stratified batches
    all_batches = []
    while len(all_batches) < num_batches:
        batch = []
        for cls, samples in list(class_groups.items()):
            # Determine the number of samples to take from this class
            count = min(
                samples_per_class_per_batch[cls],
                len(samples),
                batch_size - len(batch),
            )
            if count > 0:
                # Add samples from the current class
                batch.extend(samples[:count])
                class_groups[cls] = samples[count:]  # Remove used samples
            elif replace and len(batch) < batch_size:
                # Reuse samples if `replace` is True
                batch.extend(random.choices(samples, k=batch_size - len(batch)))

        # Shuffle the batch to mix classes
        random.shuffle(batch)
        if len(batch) == batch_size or not drop_last_batch:
            all_batches.append(batch)

    return all_batches

def get_prompt_annotations(self) -> t.Dict[str, t.List[PromptAnnotation]]:
    """
    Get all the prompt annotations for each prompt as a list.
    """
    prompt_annotations = defaultdict(list)
    for sample in self.samples:
        if sample.is_accepted:
            for prompt_name, prompt_annotation in sample.prompts.items():
                prompt_annotations[prompt_name].append(prompt_annotation)
    return prompt_annotations

def pretty_repr(self):
    """Returns a formatted string representation of the human message."""
    return f"Human: {self.content}"

def pretty_repr(self):
    """Returns a formatted string representation of the tool message."""
    return f"ToolOutput: {self.content}"

def to_dict(self, **kwargs):
    """
    Returns a dictionary representation of the AI message.
    """
    content = (
        self.content
        if self.tool_calls is None
        else {
            "text": self.content,
            "tool_calls": [tc.dict() for tc in self.tool_calls],
        }
    )
    return {"content": content, "type": self.type}

def pretty_repr(self):
    """
    Returns a formatted string representation of the AI message.
    """
    lines = []
    if self.content != "":
        lines.append(f"AI: {self.content}")
    if self.tool_calls is not None:
        lines.append("Tools:")
        for tc in self.tool_calls:
            lines.append(f"  {tc.name}: {tc.args}")

    return "\n".join(lines)