Load data from scratch

Code/python/notebooks/predict.py

@@ -31,12 +31,29 @@ warm = 900
 pred = 1600
 start_offset = 30 * 60 - warm
 end_offset = 30 * 60 - pred
-base_ds = dataloader.SaveDataset(
-        torch.load("long_dataset.pt"),
-        step=step,
-        start_offset=start_offset,
-        end_offset=end_offset,
-    )
+
+def load_file_list(path: pathlib.Path):
+    base = path.parent
+    return [
+        (base / f).resolve()
+        for f in path.read_text().splitlines()
+        if (base / f).exists()
+    ]
+
+file_list_path = pathlib.Path("../../cpp/known_routes_and_aircraft.csv")
+file_list = load_file_list(file_list_path)
+
+base_ds = dataloader.EvenlySpacedDataset(
+    filepaths=file_list,
+    n_input=30 * 60,
+    n_output=30 * 60,
+    n_windows_per_file=5,
+    step=1,
+    feature_columns=("lat", "lon", "alt", "ias"),
+    context_columns=("last_lat", "last_lon", "last_alt", "last_ias", "last_timestamp"),
+    time_columns=("timestamp", "dt"),
+    target_columns=("lat", "lon", "alt"),
+)
 train_ds, val_ds = split_dataset(base_ds)
 model = cls(
             time_in=2,
@@ -51,7 +68,7 @@ model = cls(
             device=DEVICE,
         )
 
-model.load_state_dict(torch.load("LSTM_wu900_ps150_aw0.001.pt", map_location=torch.device('cpu')))
+model.load_state_dict(torch.load("LSTM_wu900_ps150_aw0.001.pt"))
 model.to(DEVICE)
 model.eval()
 with torch.no_grad():
@@ -72,7 +89,7 @@ with torch.no_grad():
         lat_true = lat_true.flatten()
         lon_true = lon_true.flatten()
         alt_true = alt_true.flatten()
-        results.append(torch.cat([lat_true, lon_true, alt_true, lat_pred, lon_pred, alt_pred], dim=1))
+        results.append(torch.stack([lat_true, lon_true, alt_true, lat_pred, lon_pred, alt_pred], dim=1))
     
     results_tensor = torch.cat(results, dim=0)
     np_y_all = results_tensor.cpu().numpy()