add results in readme.md

README.md

@@ -60,11 +60,23 @@ All the data is publicly available and the evaluation script can be run with the
 python scripts/run_evaluation.py
 ```
 
-The test dataset we used is defined in `quartz_solar_forecast/dataset/testset.csv'. 
+The test dataset we used is defined in `quartz_solar_forecast/dataset/testset.csv`. 
 This contains 50 PV sites, which 50 unique timestamps. The data is from 2021. 
 
-The results of the evaluation are shown below:
-TODO
+The results of the evaluation are as follows
+The MAE is 0.1906 kw across all horizons. 
+
+| Horizons | MAE [kw]      |
+|----------|---------------|
+| 0        | 0.202 +- 0.03 |
+| 1        | 0.211 +- 0.03 |
+| 2        | 0.216 +- 0.03 |
+| 3-4      | 0.211 +- 0.02 |
+| 5-8      | 0.191 +- 0.01 |
+| 9-16     | 0.161 +- 0.01 |
+| 17-24    | 0.173 +- 0.01 |
+| 24-48    | 0.201 +- 0.01 |
+
 
 
 

quartz_solar_forecast/eval/metrics.py

@@ -42,4 +42,28 @@ def metrics(results_df: pd.DataFrame):
 
         print(f"MAE for horizon {horizon_hour}: {mae} +- {1.96*sem}")
 
-    # TODO add more metrics using ocf_ml_metrics
+    # calculate metrics over the different horizon groups
+    horizon_groups = [[0, 0], [1, 1], [2, 2], [3, 4], [5, 8], [9, 16], [17, 24], [24, 48]]
+    for horizon_group in horizon_groups:
+        horizon_group_df = results_df[
+            results_df["horizon_hour"].between(horizon_group[0], horizon_group[1])
+        ]
+        mae = np.round(
+            (horizon_group_df["forecast_power"] - horizon_group_df["generation_power"])
+            .abs()
+            .mean(),
+            3,
+        )
+        sem = np.round(
+            (
+                (horizon_group_df["forecast_power"] - horizon_group_df["generation_power"])
+                .abs()
+                .std()
+                / len(horizon_group_df) ** 0.5
+            ),
+            3,
+        )
+
+        print(f"MAE for horizon {horizon_group}: {mae} +- {1.96*sem}")
+
+        # TODO add more metrics using ocf_ml_metrics