Additional functions to `eurostat` Python package (https://pypi.org/project/eurostat)

eurostat_extended.py

import eurostat
import pandas as pd

def importData(EurostatDatasetCode, flags=False):
    """
    Import a dataset from Eurostat as a flat/melted table (pandas dataframe)

    Parameter
    ----------
    EurostatDatasetCode : str
        For the available datasets and their codes see:
        https://ec.europa.eu/eurostat/estat-navtree-portlet-prod/BulkDownloadListing?dir=data&start=all
    flags : bool
        Should the flags be included?

    Returns
    -------
    pandas.core.frame.DataFrame
        The Eurostat dataset as a flat/melted table.
    
    Example
    --------
    >>> importData('nama_10_gdp')
                   unit na_item geo  time     value
    0        CLV05_MEUR     B1G  AT  2021  269789.6
    1        CLV05_MEUR     B1G  BA  2021   11154.0
    2        CLV05_MEUR     B1G  BE  2021  343926.1
    3        CLV05_MEUR     B1G  BG  2021   29421.7
    4        CLV05_MEUR     B1G  CH  2021  440224.6
    ...             ...     ...  ..   ...       ...
    1363982    PYP_MNAC     YA1  PT  1975       NaN
    1363983    PYP_MNAC     YA1  RO  1975       NaN
    1363984    PYP_MNAC     YA1  SI  1975       NaN
    1363985    PYP_MNAC     YA1  UK  1975       NaN
    1363986    PYP_MNAC     YA1  XK  1975       NaN
    """
    df = eurostat.get_data_df(EurostatDatasetCode, flags)
    idx = [idx for idx, s  in enumerate(list(df.columns)) if '\\' in str(s)][0]
    if flags:
        new_col_names = [
            'value_'+x if '_value' in x else 'flag_'+x if '_flag' in x else x
            for x in df.columns
        ]
        new_col_names = [
            x.replace('_flag',"").replace('_value',"")
            for x in new_col_names
        ]
        df.columns = new_col_names
        df = pd.wide_to_long(df, stubnames=['value_','flag_'],
                             i=list(df.columns[range(idx + 1)]),
                             j='variable')
        df = df.reset_index()
        df.rename(columns = {'value_': 'value',
                             'flag_': 'flag'},
                             inplace = True)
    else:
        df = df.melt(list(df.columns[range(idx + 1)]))
    pivot = df.columns[idx]
    dim1, dim2 = pivot.split('\\')
    df.rename(columns = {pivot: dim1,
                         'variable': dim2},
                         inplace = True)
    return df