Issue/ml model (#101)

* override site model name if site.ml_model is not assigned

* update

* fix

pyproject.toml

@@ -19,7 +19,7 @@ classifiers = ["Programming Language :: Python :: 3"]
 dependencies = [
     "cryptography >= 42.0.7",
     "fastapi >= 0.105.0",
-    "pvsite-datamodel >= 1.0.41",
+    "pvsite-datamodel >= 1.0.45",
     "pyjwt >= 2.8.0",
     "pyproj >= 3.3.0",
     "pytz >= 2023.3",

src/india_api/internal/inputs/indiadb/client.py

@@ -13,6 +13,7 @@
     get_pv_generation_by_sites,
     get_user_by_email,
     get_sites_from_user,
+    get_site_by_uuid,
 )
 from pvsite_datamodel.write.generation import insert_generation_values
 from pvsite_datamodel.sqlmodels import SiteAssetType, ForecastValueSQL
@@ -103,6 +104,10 @@ def get_predicted_power_production_for_location(
 
             site = sites[0]
 
+            if site.ml_model is not None:
+                ml_model_name = site.ml_model.name
+            log.info(f"Using ml model {ml_model_name}")
+
             # read actual generations
             values = get_latest_forecast_values_by_site(
                 session,
@@ -286,6 +291,12 @@ def get_site_forecast(self, site_uuid: str, email: str) -> list[internal.Predict
         with self._get_session() as session:
             check_user_has_access_to_site(session=session, email=email, site_uuid=site_uuid)
 
+            # get site and the get the ml model name
+            site = get_site_by_uuid(session=session, site_uuid=site_uuid)
+            if site.ml_model is not None:
+                ml_model_name = site.ml_model.name
+            log.info(f"Using ml model {ml_model_name}")
+
             if isinstance(site_uuid, str):
                 site_uuid = UUID(site_uuid)
 

src/india_api/internal/inputs/indiadb/conftest.py

@@ -176,6 +176,7 @@ def make_fake_forecast_values(db_session, sites, model_name):
                     horizon_minutes=horizon,
                 )
                 forecast_value.ml_model = ml_model
+                site.ml_model = ml_model
 
                 forecast_values.append(forecast_value)