fix(deps): upgrade pandas to 3.1 and fix EOS deprecations (#1469)

* build(deps): bump pandas from 3.0.6 to 3.1.0

* fix(prediction): avoid pandas 3.1 deprecation warnings

src/akkudoktoreos/prediction/elecpriceakkudoktor.py

@@ -149,7 +149,7 @@ class ElecPriceAkkudoktor(ElecPriceProvider):
             price_wh = value.marketpriceEurocentPerKWh / (100 * 1000)
 
             # Collect all values into the Pandas Series
-            prices_wh.at[orig_datetime] = price_wh
+            prices_wh.loc[(orig_datetime,)] = price_wh
 
         if not highest_orig_datetime:  # mypy fix
             error_msg = f"Highest original datetime not available: {highest_orig_datetime}"

src/akkudoktoreos/prediction/elecpriceenergycharts.py

@@ -245,7 +245,7 @@ class ElecPriceEnergyCharts(ElecPriceProvider):
             price_wh = price_eur_per_mwh / 1_000_000
 
             # Store in series
-            prices_wh.at[orig_datetime] = price_wh
+            prices_wh.loc[(orig_datetime,)] = price_wh
 
         # Always raw here — fees are applied once, later, over the complete
         # raw+predicted series in _store_gross_series().

src/akkudoktoreos/prediction/elecpricetibber.py

@@ -301,7 +301,7 @@ class ElecPriceTibber(ElecPriceProvider):
         series_data = pd.Series(dtype=float)
         for point in points:
             orig_datetime = to_datetime(point.startsAt, in_timezone=self.config.general.timezone)
-            series_data.at[orig_datetime] = point.total / 1000.0
+            series_data.loc[(orig_datetime,)] = point.total / 1000.0
 
         return series_data.sort_index()
 

src/akkudoktoreos/prediction/feedintariffakkudoktor.py

@@ -92,7 +92,7 @@ class FeedInTariffAkkudoktor(FeedInTariffProvider):
         series = pd.Series(dtype=float)
         for value in data.values:
             timestamp = to_datetime(value.start, in_timezone=self.config.general.timezone)
-            series.at[timestamp] = value.marketprice / 1_000_000
+            series.loc[(timestamp,)] = value.marketprice / 1_000_000
         return series
 
     async def _update_data(self, force_update: Optional[bool] = False) -> None:

src/akkudoktoreos/prediction/feedintariffdvhubonline.py

@@ -111,7 +111,7 @@ class FeedInTariffDvhubOnline(FeedInTariffProvider):
             except (KeyError, TypeError, ValueError) as exc:
                 logger.warning("Skipping malformed dvhub.online price entry {}: {}", entry, exc)
                 continue
-            series_data.at[orig_datetime] = price_eur_per_mwh / 1_000_000
+            series_data.loc[(orig_datetime,)] = price_eur_per_mwh / 1_000_000
         return series_data
 
     async def _update_data(self, force_update: Optional[bool] = False) -> None:

src/akkudoktoreos/prediction/feedintariffenergycharts.py

@@ -180,7 +180,7 @@ class FeedInTariffEnergyCharts(FeedInTariffProvider):
             price_wh = price_eur_per_mwh / 1_000_000
 
             # Store in series
-            prices_wh.at[orig_datetime] = price_wh
+            prices_wh.loc[(orig_datetime,)] = price_wh
 
         # Always raw here — fees are applied once, later, over the complete
         # raw+predicted series in _store_gross_series().

src/akkudoktoreos/prediction/feedintarifftibber.py

@@ -98,7 +98,7 @@ class FeedInTariffTibber(FeedInTariffProvider):
             if point.energy is None:
                 raise ValueError("Tibber response does not contain the energy price component")
             timestamp = to_datetime(point.startsAt, in_timezone=self.config.general.timezone)
-            series.at[timestamp] = point.energy / 1000.0
+            series.loc[(timestamp,)] = point.energy / 1000.0
 
         if series.empty:
             raise ValueError("Tibber response contains no feed-in price points")

src/akkudoktoreos/server/dash/prediction.py

@@ -280,7 +280,7 @@ def Prediction(eos_host: str, eos_port: Union[str, int], data: Optional[dict] =
     # Merge into dataframe
     # ---------------------------------------------------------------------
     predictions = pd.concat(series_list, axis=1).reset_index()
-    predictions.rename(columns={"index": "date_time"}, inplace=True)
+    predictions = predictions.rename(columns={"index": "date_time"})
 
     # Remove time offset from UTC to get naive local time and make bokeh plot in local time
     date_time_tz = predictions["date_time"].dt.tz