latent.flows.validate_dataset¶

validate_dataset — Pre-validation checks for evaluation datasets.

Classes¶

ValidationResult(check_name: str, passed: bool, message: str, failed_indices: list[int] = list())

Result of a single validation check.

check_min_rows(min_count: int) -> Callable[[pd.DataFrame], ValidationResult]

Create a check that verifies minimum row count.

check_no_nulls(columns: str = ()) -> Callable[[pd.DataFrame], ValidationResult]

Create a check that verifies no null values in specified columns.

check_required_columns(columns: str = ()) -> Callable[[pd.DataFrame], ValidationResult]

Create a check that verifies required columns exist.

validate_dataset(eval_data: pd.DataFrame, checks: list[Callable[[pd.DataFrame], ValidationResult]]) -> tuple[bool, list[ValidationResult]]

Run all validation checks on a dataset.

Args: eval_data: DataFrame to validate. checks: List of callables, each taking a DataFrame and returning a ValidationResult.

Returns: Tuple of (all_passed: bool, results: list[ValidationResult]).