openavmkit.utilities._utils

to_parquet_safe

to_parquet_safe(df, path, geometry_col=None)

Writes the dataframe to parquet, performing a bunch of safety checks so you don't have to

Parameters:

Name	Type	Description	Default
df	DataFrame	Pandas dataframe or geopandas geodataframe	required
path	str	The path you want to write the parquet to	required
geometry_col	str	The name of the geometry column. Default is None	None

Source code in openavmkit/utilities/_utils.py

def to_parquet_safe(df : pd.DataFrame, path: str, geometry_col : str | None = None):
    """
    Writes the dataframe to parquet, performing a bunch of safety checks so you don't have to

    Parameters
    ----------
    df : pd.DataFrame
        Pandas dataframe or geopandas geodataframe
    path : str
        The path you want to write the parquet to
    geometry_col : str
        The name of the geometry column. Default is None
    """
    df_san = sanitize_df(df, geometry_col = geometry_col)
    if hasattr(df_san, 'to_numpy'):
        if hasattr(df_san, 'geometry'):  # GeoDataFrame-like
            df_san.to_parquet(path, engine="pyarrow")
        else:  # DataFrame-like
            df_san.to_parquet(path)
    else:
        raise TypeError("df must be a DataFrame.")