add adjuster first try

india_forecast_app/adjuster.py

@@ -0,0 +1,78 @@
+from pvsite_datamodel.sqlmodels import ForecastValueSQL, GenerationSQL, ForecastSQL
+from sqlalchemy.sql import func
+
+from datetime import datetime, timedelta
+import pandas as pd
+from typing import Optional
+
+# Wad to get all the forecast for the last 7 days made, at this time.
+# And find the ME for each forecast horizon
+
+"""
+Here is the SQL query that it based off
+
+select 
+AVG(generation.generation_power_kw - forecast_values.forecast_power_kw)
+-- generation.generation_power_kw,
+-- forecast_values.forecast_power_kw,
+-- forecast_values.start_utc,
+,horizon_minutes
+-- *
+from forecast_values 
+JOIN forecasts ON forecasts.forecast_uuid = forecast_values.forecast_uuid 
+join generation on generation.start_utc = forecast_values.start_utc
+WHERE forecast_values.start_utc >= '2024-11-05' 
+AND forecasts.site_uuid = 'adaf6be8-4e30-4c98-ac27-964447e9c8e6'
+AND generation.site_uuid = 'adaf6be8-4e30-4c98-ac27-964447e9c8e6'
+and extract(hour from forecasts.created_utc) = 7
+group by horizon_minutes
+-- order by forecast_values.start_utc
+"""
+
+
+def get_me_values(session, hour: int, start_datetime: Optional[datetime] = None) -> pd.DataFrame:
+    """
+    Get the ME values for the last 7 days for a given hour, for a given hour creation time
+
+    args:
+    hour: the hour of whent he forecast is created
+    start_datetime:: the start datetime to filter on.
+    session: sqlalchemy session
+
+    """
+
+    if start_datetime is None:
+        start_datetime = datetime.now() - timedelta(days=7)
+
+    query = session.query(
+        func.avg(ForecastValueSQL.forecast_power_kw - GenerationSQL.generation_power_kw),
+        ForecastValueSQL.horizon_minutes,
+    )
+
+    # join
+    query = query.join(ForecastSQL)
+    query = query.filter(GenerationSQL.start_utc == ForecastValueSQL.start_utc)
+
+    # only include the last 7 days
+    query = query.filter(ForecastValueSQL.start_utc >= start_datetime)
+    query = query.filter(GenerationSQL.start_utc >= start_datetime)
+
+    # filter on site
+    query = query.filter(ForecastSQL.site_uuid == "adaf6be8-4e30-4c98-ac27-964447e9c8e6")
+    query = query.filter(GenerationSQL.site_uuid == "adaf6be8-4e30-4c98-ac27-964447e9c8e6")
+
+    # filter on created_utc
+    query = query.filter((func.extract("hour", ForecastSQL.created_utc) == hour))
+
+    # group by forecast horizon
+    query = query.group_by(ForecastValueSQL.horizon_minutes)
+
+    # order by forecast horizon
+    query = query.order_by(ForecastValueSQL.horizon_minutes)
+
+    me = query.all()
+
+    me_df = pd.DataFrame(me, columns=["me_kw", "horizon_minutes"])
+
+    return me_df
+

india_forecast_app/app.py

@@ -21,6 +21,7 @@
 import india_forecast_app
 from india_forecast_app.models import PVNetModel, get_all_models
 from india_forecast_app.sentry import traces_sampler
+from india_forecast_app.adjuster import get_me_values
 
 log = logging.getLogger(__name__)
 version = india_forecast_app.__version__
@@ -204,6 +205,7 @@ def save_forecast(
     write_to_db: bool,
     ml_model_name: Optional[str] = None,
     ml_model_version: Optional[str] = None,
+    use_adjuster: bool = True,
 ):
     """
     Saves a forecast for a given site & timestamp
@@ -238,6 +240,32 @@ def save_forecast(
             ml_model_version=ml_model_version,
         )
 
+    if use_adjuster:
+        log.info(f"Adjusting forecast for site_id={forecast_meta['site_uuid']}...")
+        me_values = get_me_values(db_session, forecast_meta["timestamp_utc"].hour)
+
+        log.info(f"ME values: {me_values}")
+
+        # TODO
+        # should we put some sort of limits on the adjuster
+        # perhaps 10% of capacity?
+
+        # join forecast_values_df with me_values on horizon_minutes
+        forecast_values_df_adjust = forecast_values_df.copy()
+        forecast_values_df_adjust = forecast_values_df_adjust.merge(me_values, on="horizon_minutes", how="left")
+
+        # adjust forecast_power_kw by ME values
+        forecast_values_df_adjust["forecast_power_kw"] = forecast_values_df_adjust["forecast_power_kw"] - forecast_values_df_adjust["me_kw"]
+
+        if write_to_db:
+            insert_forecast_values(
+                db_session,
+                forecast_meta,
+                forecast_values_df_adjust,
+                ml_model_name=f'{ml_model_name}_adjust',
+                ml_model_version=ml_model_version,
+            )
+
     output = f'Forecast for site_id={forecast_meta["site_uuid"]},\
                timestamp={forecast_meta["timestamp_utc"]},\
                version={forecast_meta["forecast_version"]}:'