Fix deprecations in pandas 3

Author: jnettels

lpagg/BDEW.py

@@ -38,6 +38,7 @@
 import numpy as np
 import pandas as pd
 from pandas.tseries.frequencies import to_offset
+from pandas.api.types import is_numeric_dtype
 import logging
 import importlib.resources
 
@@ -91,9 +92,18 @@ def load_BDEW_style_profiles(source_file, weather_data, cfg, houses_dict,
                                   skiprows=[0], header=[0, 1], index_col=[0],
                                   skipfooter=1,
                                   )
-    weather_daily = (weather_data.resample('D', label='right', closed='right')
-                     .mean(numeric_only=True))
-    # print(weather_daily)
+
+    agg_dict = dict()
+    for col in weather_data.columns:
+        if is_numeric_dtype(weather_data[col]):
+            agg_dict[col] = 'mean'
+        else:
+            agg_dict[col] = 'first'
+
+    weather_daily = (weather_data
+                     .shift(periods=-1, freq="infer")
+                     .resample('D', label='left', closed='left')
+                     .agg(agg_dict))
 
     houses_list = settings['houses_list_BDEW']
     multiindex = pd.MultiIndex.from_product([houses_list, [energy_type]],
@@ -127,12 +137,8 @@ def load_BDEW_style_profiles(source_file, weather_data, cfg, houses_dict,
         profile_year = pd.Series(dtype='float')
         profiles_daily = []
         for date in weather_daily.index:
-            weekday = weather_data.loc[date]['weekday_BDEW']
-            season = weather_data.loc[date]['season_BDEW']
-            # Important: In order identify the weekday of the resampled days,
-            # we labled them 'right'. From now on we need the label 'left',
-            # so we substract '1 day' from each date:
-            date -= pd.Timedelta('1 day')
+            weekday = weather_daily.loc[date]['weekday_BDEW']
+            season = weather_daily.loc[date]['season_BDEW']
 
             source_profile = source_df[house_type][season, weekday]
 

lpagg/agg.py

@@ -1281,8 +1281,8 @@ def apply_DST(df, tz_default='Etc/GMT-1', tz_DST='CET', normalize=True):
     df_DST = df_DST[df_DST.index.notnull()]
     # In October, the repeated DST-hour is not in the index yet.
     # Reindex with the initial index. This will cause missing rows in October.
-    # We fill them with method forward fill.
-    df_DST = df_DST.reindex_like(df_default, method='ffill')
+    # We fill them with the 'forward fill' method.
+    df_DST = df_DST.reindex_like(df_default).ffill()
     # Finally, make the DataFrame timezone-unaware before returning it.
     df_DST = df_DST.tz_localize(None)
 

lpagg/simultaneity.py

@@ -246,7 +246,7 @@ def copy_df_columns(df, copies_list, level=None):
             else:
                 df_tmp = pd.concat([df[col]], keys=[copy_name],
                                    names=[level], axis=1)
-            df_new = pd.concat([df_new, df_tmp], axis='columns')
+            df_new = pd.concat([df_new, df_tmp], axis='columns', sort=False)
     return df_new
 
 
@@ -705,7 +705,7 @@ def plot_shifted_lineplots(df_shift, df_ref, cfg):
 
         # Save raw data
         df_excel = pd.concat([load_shift, load_ref], axis='columns',
-                             keys=[txt_shift, txt_ref])
+                             keys=[txt_shift, txt_ref], sort=False)
         df_excel.to_excel(os.path.join(cfg['print_folder'],
                                        ylabel + ' (shift)'+'.xlsx'))