Add option to use t0 embedding features

configs.example/datamodule/configuration/example_configuration.yaml

@@ -273,3 +273,9 @@ input_data:
     interval_start_minutes: -60
     interval_end_minutes: 480
     time_resolution_minutes: 30
+
+  # Note that each "cyclic" embedding adds 2 elements to the output vector to embed a period whilst 
+  # "linear" adds only 1 element. This needs to be matched to the model `t0_embedding_dim` param. 
+  # e.g. these embedding would require t0_embedding_dim = 3 (1 + 2).
+  t0_embedding:
+    embeddings: [["1h", "linear"], ["24h", "cyclic"]]

configs.example/model/late_fusion.yaml

@@ -69,6 +69,7 @@ model:
   embedding_dim: 16
   include_sun: True
   include_generation_history: False
+  t0_embedding_dim: 3
 
   # The mapping between the location IDs and their embedding indices
   location_id_mapping:

pvnet/models/late_fusion/late_fusion.py

@@ -46,6 +46,7 @@ def __init__(
         include_generation_history: bool = False,
         include_sun: bool = True,
         include_time: bool = False,
+        t0_embedding_dim: int = 0,
         location_id_mapping: dict[Any, int] | None = None,
         embedding_dim: int = 16,
         forecast_minutes: int = 30,
@@ -85,6 +86,8 @@ def __init__(
             include_generation_history: Include generation yield data.
             include_sun: Include sun azimuth and altitude data.
             include_time: Include sine and cosine of dates and times.
+            t0_embedding_dim: Shape of the embedding of the init-time (t0) of the forecast. Not used
+                if set to 0.
             location_id_mapping: A dictionary mapping the location ID to an integer. ID embedding is
                 not used if this is not provided.
             embedding_dim: Number of embedding dimensions to use for location ID.
@@ -119,6 +122,7 @@ def __init__(
         self.include_pv = pv_encoder is not None
         self.include_sun = include_sun
         self.include_time = include_time
+        self.t0_embedding_dim = t0_embedding_dim
         self.location_id_mapping = location_id_mapping
         self.embedding_dim = embedding_dim
         self.add_image_embedding_channel = add_image_embedding_channel
@@ -246,6 +250,8 @@ def __init__(
             # Update num features
             fusion_input_features += 32
 
+        fusion_input_features += self.t0_embedding_dim
+
         if include_generation_history:
             # Update num features
             fusion_input_features += self.history_len + 1
@@ -321,6 +327,9 @@ def forward(self, x: TensorBatch) -> torch.Tensor:
             time = self.time_fc1(time)
             modes["time"] = time
 
+        if self.t0_embedding_dim>0:
+            modes["t0_embed"] = x["t0_embedding"]
+
         out = self.output_network(modes)
 
         if self.use_quantile_regression:

pyproject.toml

@@ -12,7 +12,7 @@ readme = {file="README.md", content-type="text/markdown"}
 requires-python = ">=3.11,<3.14"
 
 dependencies = [
-    "ocf-data-sampler>=0.6.0",
+    "ocf-data-sampler>=1.0.9",
     "numpy",
     "pandas",
     "matplotlib",

tests/conftest.py

@@ -327,6 +327,7 @@ def raw_late_fusion_model_kwargs_generation_history(model_minutes_kwargs) -> dic
         embedding_dim=None,
         include_sun=False,
         include_time=True,
+        t0_embedding_dim=3,
         include_generation_history=True,
         forecast_minutes=480, 
         history_minutes=60,

tests/test_data/data_config.yaml

@@ -125,3 +125,6 @@ input_data:
     interval_start_minutes: -60
     interval_end_minutes: 480
     time_resolution_minutes: 30
+
+  t0_embedding:
+    embeddings: [["1h", "linear"], ["24h", "cyclic"]]