data/climatic_data.py

import openmeteo_requests
import pandas as pd
from retry_requests import retry
import pvlib
import requests_cache


def get_tmy(latitude: float, longitude: float) -> pd.DataFrame:
    tmy = pvlib.iotools.get_pvgis_tmy(
        latitude,
        longitude,
        outputformat="json",
        usehorizon=True,
        userhorizon=None,
        startyear=2005,
        endyear=2020,
        map_variables=True,
        url="https://re.jrc.ec.europa.eu/api/v5_2/",
        timeout=60,
    )[0]

    daily_tmy = tmy.groupby(tmy.index.date).agg(
        {"temp_air": ["min", "max", "mean"], "wind_speed": "mean", "ghi": "sum"}
    )
    daily_tmy.columns = daily_tmy.columns.droplevel(0)
    daily_tmy.columns = ["temp_min", "temp_max", "temp_mean", "wind", "irradiance"]

    # On récupère les données de précipitations d'Open-Meteo car elles ne sont pas inclues dans la tmy
    # Setup the Open-Meteo API client with cache and retry on error
    # TODO : utiliser package météo
    cache_session = requests_cache.CachedSession(".cache", expire_after=-1)
    retry_session = retry(cache_session, retries=5, backoff_factor=0.2)
    openmeteo = openmeteo_requests.Client(session=retry_session)

    url = "https://archive-api.open-meteo.com/v1/archive"
    params = {
        "latitude": latitude,
        "longitude": longitude,
        "start_date": "2005-01-01",
        "end_date": "2022-01-01",
        "daily": "precipitation_sum",
    }
    responses = openmeteo.weather_api(url, params=params)

    response = responses[0]
    daily = response.Daily()
    precipitation_sum = daily.Variables(0).ValuesAsNumpy()

    daily_rain = {
        "date": pd.date_range(
            start=pd.to_datetime(daily.Time(), unit="s"),
            end=pd.to_datetime(daily.TimeEnd(), unit="s"),
            freq=pd.Timedelta(seconds=daily.Interval()),
            inclusive="left",
        )
    }
    daily_rain["rain"] = precipitation_sum

    daily_rain = pd.DataFrame(data=daily_rain)
    daily_rain.date = pd.to_datetime(daily_rain.date)
    daily_rain = daily_rain.set_index("date")

    # on sélectionne les mois de daily_rain qui sont les mois dans la tmy de pvlib
    df = daily_tmy.join(daily_rain)  # c'est là qu'on garde que les mois d'intérêt
    df.index = df.index.map(lambda t: t.replace(year=1990))

    return df