add adjuster for gsp_id=0

src/gsp.py

@@ -106,6 +106,15 @@ def get_all_available_forecasts(
             f"The option is {historic=} for user {user}"
         )
 
+        # adjust gsp_id 0
+        if 0 in gsp_ids:
+            idx = [i for i, forecasts in enumerate(forecasts.forecasts) if forecasts.location.gsp_id == 0]
+            if len(idx) > 0:
+                logger.info(f"Adjusting forecast values for gsp id 0, {adjust_limit}")
+                forecasts.forecasts[idx[0]] = forecasts.forecasts[idx[0]].adjust(limit=adjust_limit)
+            else:
+                logger.warning(f"Could not find gsp id 0 in the forecasts")
+
     return forecasts
 
 

src/pydantic_models.py

@@ -1,5 +1,6 @@
 """ pydantic models for API"""
 import logging
+import os
 from datetime import datetime
 from typing import Dict, List, Optional
 
@@ -10,6 +11,9 @@
 logger = logging.getLogger(__name__)
 
 
+adjust_limit = float(os.getenv("ADJUST_MW_LIMIT", 0.0))
+
+
 class GSPYield(EnhancedBaseModel):
     """GSP Yield data"""
 
@@ -146,7 +150,23 @@ def convert_forecasts_to_many_datetime_many_generation(
                 continue
             if end_datetime_utc is not None and datetime_utc > end_datetime_utc:
                 continue
-            forecast_mw = round(forecast_value.expected_power_generation_megawatts, 2)
+
+            forecast_mw = forecast_value.expected_power_generation_megawatts
+
+            # adjust the value if gsp id 0, this is the national
+            if gsp_id == '0':
+                adjust_mw = forecast_value.adjust_mw
+                if adjust_mw > adjust_limit:
+                    adjust_mw = adjust_limit
+                elif adjust_mw < -adjust_limit:
+                    adjust_mw = -adjust_limit
+
+                forecast_mw = forecast_mw - adjust_mw
+
+                if forecast_mw < 0:
+                    forecast_mw = 0.0
+
+            forecast_mw = round(forecast_mw, 2)
 
             # if the datetime object is not in the dictionary, add it
             if datetime_utc not in many_forecast_values_by_datetime: