# Copyright (c) 2021. Bram Vonk, Enexis import logging import pandas as pd [docs]logger = logging.getLogger("SPARK") [docs]def asses_forecasts(df_total, df_meta): """ Assesses if and when a transformer will overload based on the forecasts and metadata. Parameters ---------- df_total: pd.DataFrame DataFrame with forecasts. df_meta: pd.DataFrame DataFrame with metadata of the transformer. Returns ------- pd.DataFrame Result of the assessment. """ logger.info("Assessing forecast") # get maximum of forecast df_assess = df_total.query("period == 'future' & model_var == 'Σ'").copy() df_assess["relative_abs_max"] = (df_assess["value"].abs() / df_meta["vermogen_nominaal"].squeeze()).clip(upper=1) df_assess["date_abs_max"] = df_assess["date"] df_assess = df_assess.sort_values(by=["relative_abs_max", "date_abs_max"], ascending=[False, True]).iloc[[0]] df_forecast_meta = pd.merge(df_meta, df_assess, on="boxid") df_forecast_meta["regio"] = df_forecast_meta["regio_netautomatisering"] cols = [ "boxid", "stationsnummer", "regio", "vestiging", "l", "processed_on", "vermogen_nominaal", "relative_abs_max", "date_abs_max", ] return df_forecast_meta[cols]