instancespace.data.metadata

Defines data types for metadata.

These classes define types for problem instances found in the metadata.csv file.

MATRIX_DIMENSIONS = 2

@dataclass(frozen=True)

class Metadata:

Metadata for problem instances.

Metadata( feature_names: list[str], algorithm_names: list[str], instance_labels: pandas.core.series.Series, instance_sources: pandas.core.series.Series | None, features: numpy.ndarray[typing.Any, numpy.dtype[numpy.float64]], algorithms: numpy.ndarray[typing.Any, numpy.dtype[numpy.float64]])

feature_names: list[str]

algorithm_names: list[str]

instance_labels: pandas.core.series.Series

instance_sources: pandas.core.series.Series | None

features: numpy.ndarray[typing.Any, numpy.dtype[numpy.float64]]

algorithms: numpy.ndarray[typing.Any, numpy.dtype[numpy.float64]]

@staticmethod

def from_data_frame( data: pandas.core.frame.DataFrame) -> Metadata:

Parse metadata from a file, and construct a Metadata object.

Parameters

data: The content of a csv file containing the metadata.

Returns

Metadata: A Metadata object.

def from_csv_file( file_path: pathlib.Path | str) -> Metadata | None:

Parse metadata from a CSV file and construct a Metadata object.

Parameters

file_path (Path | str): The path to the CSV file containing the metadata.

Returns

Metadata or None: A Metadata object constructed from the parsed CSV data, or None if an error occurred during file reading or parsing.

Raises

FileNotFoundError: If the specified file does not exist.
pandas.errors.EmptyDataError: If the specified file is empty.
pandas.errors.ParserError: If the specified file is not a valid CSV file.