Merge pull request #144 from openclimatefix/jacob/persistence

Add Persistence Baseline

pvnet/models/base_model.py

@@ -471,11 +471,23 @@ def validation_step(self, batch: dict, batch_idx):
         y_hat = self(batch)
         # Sensor seems to be in batch, station, time order
         y = batch[self._target_key][:, -self.forecast_len :, 0]
-
+        persistence = batch[self._target_key][:, -self.forecast_len - 1, 0]
+        # Expand persistence to be the same shape as y
+        persistence = persistence.unsqueeze(1).expand(-1, self.forecast_len)
         losses = self._calculate_common_losses(y, y_hat)
         losses.update(self._calculate_val_losses(y, y_hat))
 
         logged_losses = {f"{k}/val": v for k, v in losses.items()}
+        logged_losses["MAE_persistence/val"] = F.l1_loss(persistence, y)
+        logged_losses["MSE_persistence/val"] = F.mse_loss(persistence, y)
+        # Log for each timestep the persistence loss
+        for i in range(self.forecast_len):
+            logged_losses[f"MAE_persistence/step_{i:03}/val"] = F.l1_loss(
+                persistence[:, i], y[:, i]
+            )
+            logged_losses[f"MSE_persistence/step_{i:03}/val"] = F.mse_loss(
+                persistence[:, i], y[:, i]
+            )
         # Get the losses in the format of {VAL>_horizon/step_000: 0.1, VAL>_horizon/step_001: 0.2}
         # for each step in the forecast horizon
         # This is needed for the custom plot