domainiac 9.2.0__py3-none-any.whl → 9.3.0__py3-none-any.whl

domainiac/managers/nwp_manager.py

@@ -1,3 +1,5 @@
+from functools import lru_cache
+
 import datamazing.pandas as pdz
 import numpy as np
 import pandas as pd
@@ -9,26 +11,36 @@ from ..modeling.nwp import Coordinate, Neighborhood, NWPParameter, NWPProvider
 
 
 class NWPManager:
-    """
-    Contains all logic concerning NWP data.
-    This includes:
-    - Getting NWP data
-    - Getting average NWP data
-    - Finding closest NWP coordinates
-    """
-
     def __init__(
         self,
         db: pdz.Database,
         time_interval: pdz.TimeInterval,
         resolution: pd.Timedelta,
+        cache_nwp: bool = True,
     ):
+        """
+        Contains all logic concerning NWP data.
+        This includes:
+        - Getting NWP data
+        - Getting average NWP data
+        - Finding closest NWP coordinates
+
+        Args:
+            db (pdz.Database): Database backend
+            time_interval (pdz.TimeInterval): Time interval
+            resolution (pd.Timedelta): Time resolution
+            cache_nwp (bool, optional): Current implementation of this managers
+                is not very efficient when querying multiple coordinates. To mitigate
+                this, database queries can be cached. This is especially useful when
+                querying multiple coordinates. Defaults to True.
+        """
         self.db = db
         self.time_interval = time_interval
         self.resolution = resolution
 
         self._nwp_coordinates_kd_tree = dict()
         self._nwp_table_name_prefix = "forecastNwp"
+        self.cache_nwp = cache_nwp
 
     @staticmethod
     def calculate_wind_speed_from_vectors(u: pd.Series, v: pd.Series) -> pd.Series:
@@ -111,15 +123,24 @@ class NWPManager:
 
         return coordinates
 
+    @lru_cache()
+    def _query_cached(
+        self,
+        table: str,
+        time_interval: pdz.TimeInterval,
+    ) -> pd.DataFrame:
+        """
+        Query the database with caching.
+        """
+        df = self.db.query(table, time_interval)
+        return df
+
     def _get_nwp_parameter(
         self,
         provider: NWPProvider,
         parameter: NWPParameter,
         coordinate: Coordinate,
     ) -> pd.DataFrame:
-        """
-        Get NWP data for a given parameter, coordinate and altitude.
-        """
         table = self._get_table_name(provider, parameter)
 
         filters = {
@@ -138,7 +159,13 @@ class NWPManager:
             self.time_interval.right + query_margin,
         )
 
-        df = self.db.query(table, padded_time_interval, filters=filters)
+        # Cache query if specified
+        # This is useful when querying multiple coordinates.
+        if self.cache_nwp:
+            df = self._query_cached(table, padded_time_interval)
+            df = df.loc[(df[list(filters)] == pd.Series(filters)).all(axis=1)]
+        else:
+            df = self.db.query(table, padded_time_interval, filters=filters)
 
         df = df.drop(
             columns=[
@@ -160,10 +187,6 @@ class NWPManager:
         parameter: NWPParameter,
         coordinate: Coordinate,
     ):
-        """
-        Get NWP data for a given parameter, coordinate and altitude,
-        using the proper interpolation technique.
-        """
         df = self._get_nwp_parameter(provider, parameter, coordinate)
 
         df = df.sort_values(by="time_utc")
@@ -214,9 +237,12 @@ class NWPManager:
         parameter: NWPParameter,
         coordinate: Coordinate,
     ):
-        df_nwp = self._get_interpolated_nwp_parameter(provider, parameter, coordinate)
+        df_nwp = self._get_interpolated_nwp_parameter(
+            provider,
+            parameter,
+            coordinate,
+        )
 
-        # Ensure data is within time_interval defined by user
         df_nwp = df_nwp.query(
             f"time_utc >= '{self.time_interval.left}' "
             f"and time_utc <= '{self.time_interval.right}'",
@@ -230,14 +256,15 @@ class NWPManager:
         parameter: NWPParameter,
         neighborhood: Neighborhood,
     ) -> pd.DataFrame:
-        """
-        Get average NWP data for a given parameter, coordinates and altitude.
-        """
         dfs = []
 
         neighbors = self.get_nwp_neighbors(provider, parameter, neighborhood)
         for coordinate in neighbors:
-            df_coordinate = self.get_nwp_parameter(provider, parameter, coordinate)
+            df_coordinate = self.get_nwp_parameter(
+                provider,
+                parameter,
+                coordinate,
+            )
             dfs.append(df_coordinate)
 
         df = pd.concat(dfs)
@@ -252,7 +279,19 @@ class NWPManager:
         neighborhood: Neighborhood,
     ) -> pd.DataFrame:
         """
-        Get weather features for a given coordinate.
+        Get NWP data for multiple parameters over a neighborhood.
+
+        For each parameter, retrieves the average NWP data in the neighborhood, then
+        merges all results on the time_utc column.
+
+        Args:
+            provider (NWPProvider): The NWP data provider to use.
+            parameters (list[NWPParameter]): List of NWP parameters to retrieve.
+            neighborhood (Neighborhood): The neighborhood specifying the central
+                coordinate and number of neighbors.
+
+        Returns:
+            pd.DataFrame: DataFrame with merged parameter values, indexed by time_utc.
         """
         dfs = []
         for parameter in parameters:

{domainiac-9.2.0.dist-info → domainiac-9.3.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: domainiac
-Version: 9.2.0
+Version: 9.3.0
 Summary: Package for working with Energinet data, but with specialized functions used for Enigma.
 Author: Team Enigma
 Author-email: enigma@energinet.dk

{domainiac-9.2.0.dist-info → domainiac-9.3.0.dist-info}/RECORD

@@ -9,7 +9,7 @@ domainiac/functions/wind.py,sha256=ZOxSMxjWV7LZsAQVW5Y-GPMhZzfvIsh3NOtSXB5lUAY,9
 domainiac/managers/__init__.py,sha256=7N-iEsFidupPi1u3EURfeWAoQ-NdrzAwR13faA_VWUs,298
 domainiac/managers/masterdata_manager.py,sha256=_xIHEWVYssdvC2VMujvR0ANXCI_Er3qOlfAX427yimk,2298
 domainiac/managers/metering_manager.py,sha256=MauMncY_yaVPx8Y9CCD8nJiE1BGHFKTVC1bOtiyl2pI,1701
-domainiac/managers/nwp_manager.py,sha256=rEN9HmDe01N5QzePGsUorWgmOuY7t9vMwBF1Kuuu_38,8603
+domainiac/managers/nwp_manager.py,sha256=h2J6ryXqplB_W6jC7aSenOa9MOxsYrd2HenU3WZnyu0,10040
 domainiac/managers/outage_manager.py,sha256=rH4NYY0hiD51V6SadVgtIOX0qCjG2cSZ60F9lDjkJ8o,1416
 domainiac/managers/plant_manager.py,sha256=aqJ04QB-QI83zIllQiyTrlh_SKR6CYI4bY4z4RlBL3g,5083
 domainiac/managers/resource_manager.py,sha256=QMkNcrgyXLLlkCR9kg4-nFjTaYHtlsldex-b_8lAmO0,3730
@@ -19,6 +19,6 @@ domainiac/modeling/nwp.py,sha256=PolrBdQn8W-e1M0_pefvmLn2mr4HT0NqlYlkyCV0dds,438
 domainiac/modeling/plant.py,sha256=Y0_Q6V6Lj3irJh5z-49r5gFJdD4Xl_pfHIMUhdaJVUI,2226
 domainiac/wrappers/__init__.py,sha256=vZOw9maXgVoAvudZqioD-GTPkgPI6fm7_CUELQcR_-g,43
 domainiac/wrappers/cache_wrapper.py,sha256=jDg-Lt_y-YzItyP-74tGULOxb07s_CcutmOHP1Ie00Q,355
-domainiac-9.2.0.dist-info/METADATA,sha256=_07orTSIon0JOGS6raqGn476L0GGh65-uNd8cksbx7g,737
-domainiac-9.2.0.dist-info/WHEEL,sha256=zp0Cn7JsFoX2ATtOhtaFYIiE2rmFAD4OcMhtUki8W3U,88
-domainiac-9.2.0.dist-info/RECORD,,
+domainiac-9.3.0.dist-info/METADATA,sha256=tFNeV_xnaOi5u0GcSRaCyZkdhL0oHH87X3JgbW2W8Nw,737
+domainiac-9.3.0.dist-info/WHEEL,sha256=zp0Cn7JsFoX2ATtOhtaFYIiE2rmFAD4OcMhtUki8W3U,88
+domainiac-9.3.0.dist-info/RECORD,,