elec_dt_parser.py

import pandas as pd

def dt_rng_to_SPs(start_date:datetime, end_date:datetime, freq='30T', tz='Europe/London'):
    dt_rng = pd.date_range(start_date, end_date, freq=freq, tz=tz)
    dt_strs = dt_rng.strftime('%Y-%m-%d')

    dt_SP_counts = pd.Series(dt_strs).groupby(dt_strs).count()
    SPs = []

    for num_SPs in dt_SP_counts.values:
        SPs += [SP+1 for SP in list(range(num_SPs))]

    df_dates_SPs = pd.DataFrame({'date':dt_strs, 'SP':SPs}, index=dt_rng)

    return df_dates_SPs

def assign_local_datetime(df, dt_col='settlementDate', SP_col='settlementPeriod', freq='30T', tz='Europe/London'):
    start_date = df[dt_col].min()
    end_date = df[dt_col].max()

    start_date = (pd.to_datetime(start_date) - pd.Timedelta(days=1)).strftime('%Y-%m-%d')
    end_date = (pd.to_datetime(end_date) + pd.Timedelta(days=1)).strftime('%Y-%m-%d')

    ts_to_dt_SPs = dt_rng_to_SPs(start_date, end_date, freq=freq, tz=tz)

    date_SP_tuples = list(zip(ts_to_dt_SPs['date'], ts_to_dt_SPs['SP']))
    dt_SP_to_ts = dict(zip(date_SP_tuples, ts_to_dt_SPs.index))

    s_dt_SPs = pd.Series(zip(df[dt_col], df[SP_col].astype(int)), index=df.index)

    df = df.assign(datetime=s_dt_SPs.map(dt_SP_to_ts))
    
    return df