Copernicus - FWI module¶
-
class
pyro_risks.datasets.fwi.GwisFwi(days_list: Optional[List[str]] = None)[source]¶ Bases:
pandas.core.frame.DataFrameGWIS FWI dataframe (8 km resolution) on French territory based on 2019-2020 data.
-
pyro_risks.datasets.fwi.create_departement_df(day_data: pandas.core.frame.DataFrame) → pandas.core.frame.DataFrame[source]¶ Create dataframe with lon, lat coordinates and corresponding departments.
Load json with the department polygons and run function include_department to get the name of departments corresponding to each row of input data, typically one day of FWI data got with load_data. This may take a few minutes due to the shapely process. :param day_data: df with longitudes and latitudes :type day_data: pd.Dataframe
- Returns
dataframe with lat, lon and department
- Return type
pd.DataFrame
-
pyro_risks.datasets.fwi.get_fwi_data(source_path: str, day: Optional[str] = '20190101') → pandas.core.frame.DataFrame[source]¶ Load and handle netcdf data for selected day.
Return pandas dataframe with longitude, latitude, day and fwi indices (fwi, ffmc, dmc, dc, isi, bui, dsr, dr). :param source_path: path with unzipped netcdf fwi data, usually got from load_data. :type source_path: str :param day: which day to load. Defaults to ‘20190101’. :type day: str, optional
- Returns
dataframe with all fwi indices for selected day
- Return type
pd.DataFrame
-
pyro_risks.datasets.fwi.get_fwi_data_for_predict(date: str) → pandas.core.frame.DataFrame[source]¶ Run CDS API queries for dates required by the model and return fwi dataset for predict step.
This takes care principally of the lags required for the modelling step.
- Parameters
date (str) –
- Returns
pd.DataFrame
-
pyro_risks.datasets.fwi.get_fwi_from_api(date: str) → geopandas.geodataframe.GeoDataFrame[source]¶ Call the CDS API and return all fwi variables as a dataframe with geo coordinates and departments.
When calling the API we get a zip file that must be extracted (in a tmp directory), then handle each queried variable which is in a separate netcdf file. A dataframe is created with all the variables and then finally we join codes and departments with geopandas.
- Parameters
date (str) –
- Returns
pd.DataFrame
-
pyro_risks.datasets.fwi.include_department(row: pandas.core.series.Series, polygons_json: Dict[str, Any]) → str[source]¶ Given a row of a dataframe containing longitude and latitude returns name of french department.
This function makes use of shapely to return if a polygon contains a point. :param row: row of dataframe :type row: pd.Series :param polygons_json: dict with polygons of the departments :type polygons_json: dict
- Returns
name of department or empty string
- Return type
str