From 75ef0a5b312a79b97a671ae071f8bb28304250e5 Mon Sep 17 00:00:00 2001
From: Peter Dudfield <34686298+peterdudfield@users.noreply.github.com>
Date: Fri, 4 Oct 2024 16:50:08 +0100
Subject: [PATCH] Large capacity (#205)

* TDD: update test first

* make forecast work for large sites

* round results in test
---
 quartz_solar_forecast/forecast.py | 16 ++++++++++++++++
 tests/test_forecast.py            | 20 +++++++++++++++++++-
 2 files changed, 35 insertions(+), 1 deletion(-)

diff --git a/quartz_solar_forecast/forecast.py b/quartz_solar_forecast/forecast.py
index 7cc7f2d..a75647b 100644
--- a/quartz_solar_forecast/forecast.py
+++ b/quartz_solar_forecast/forecast.py
@@ -1,4 +1,5 @@
 from datetime import datetime, timedelta
+import logging
 
 import pandas as pd
 
@@ -6,6 +7,7 @@
 from quartz_solar_forecast.forecasts import forecast_v1_tilt_orientation, TryolabsSolarPowerPredictor
 from quartz_solar_forecast.pydantic_models import PVSite
 
+log = logging.getLogger(__name__)
 
 def predict_ocf(
     site: PVSite, model=None, ts: datetime | str = None, nwp_source: str = "icon"
@@ -25,6 +27,16 @@ def predict_ocf(
     if isinstance(ts, str):
         ts = datetime.fromisoformat(ts)
 
+    if site.capacity_kwp > 4:
+        log.warning("Your site capacity is greater than 4kWp, "
+                    "however the model is trained on sites with capacity <= 4kWp."
+                    "We therefore will run the model with a capacity of 4 kWp, "
+                    "and we'll scale the results afterwards.")
+        capacity_kwp_original = site.capacity_kwp
+        site.capacity_kwp = 4
+    else:
+        capacity_kwp_original = site.capacity_kwp
+
     # make pv and nwp data from nwp_source
     nwp_xr = get_nwp(site=site, ts=ts, nwp_source=nwp_source)
     pv_xr = make_pv_data(site=site, ts=ts)
@@ -32,6 +44,10 @@ def predict_ocf(
     # load and run models
     pred_df = forecast_v1_tilt_orientation(nwp_source, nwp_xr, pv_xr, ts, model=model)
 
+    # scale the results if the capacity is different
+    if capacity_kwp_original != site.capacity_kwp:
+        pred_df["power_kw"] = pred_df["power_kw"] * capacity_kwp_original / site.capacity_kwp
+
     return pred_df
 
 
diff --git a/tests/test_forecast.py b/tests/test_forecast.py
index 975103a..c2d0c70 100644
--- a/tests/test_forecast.py
+++ b/tests/test_forecast.py
@@ -2,6 +2,9 @@
 from quartz_solar_forecast.pydantic_models import PVSite
 from datetime import datetime, timedelta
 
+import numpy as np
+
+
 def test_run_forecast():
     # make input data
     site = PVSite(latitude=51.75, longitude=-1.25, capacity_kwp=1.25)
@@ -55,7 +58,22 @@ def test_run_forecast_historical():
     print("\n Prediction based on UKMO NWP\n")
     print(predications_df_ukmo)
     print(f" Max: {predications_df_ukmo['power_kw'].max()}")
-    
+
     print("\n Prediction based on XGB\n")
     print(predications_df_xgb)
 
+
+def test_large_capacity():
+
+    # make input data
+    site = PVSite(latitude=51.75, longitude=-1.25, capacity_kwp=4)
+    site_large = PVSite(latitude=51.75, longitude=-1.25, capacity_kwp=4000)
+    ts = datetime.today() - timedelta(weeks=2)
+
+    # run model with icon, gfs and ukmo nwp
+    predications_df = run_forecast(site=site, model="gb", ts=ts, nwp_source="gfs")
+    predications_df_large = run_forecast(site=site_large, model="gb", ts=ts, nwp_source="gfs")
+
+    assert np.round(predications_df["power_kw"].sum() * 1000, 8) == np.round(
+        predications_df_large["power_kw"].sum(), 8
+    )