Branch ci lint typecheck fix

.github/workflows/branch_ci.yml

@@ -15,5 +15,4 @@ jobs:
       enable_typechecking: false
       containerfile: 'None'
       tests_folder: 'tests'
-      tests_matrix: true
       test_python_versions: '["3.11", "3.12"]'

pvnet/datamodule.py

@@ -60,18 +60,18 @@ def __init__(
         self.seed = seed
         self.dataset_pickle_dir = dataset_pickle_dir
 
-        self._common_dataloader_kwargs = dict(
-            batch_size=batch_size,
-            batch_sampler=None,
-            num_workers=num_workers,
-            collate_fn=collate_fn,
-            pin_memory=pin_memory,
-            timeout=0,
-            worker_init_fn=None,
-            prefetch_factor=prefetch_factor,
-            persistent_workers=persistent_workers,
-            multiprocessing_context="spawn" if num_workers > 0 else None,
-        )
+        self._common_dataloader_kwargs = {
+            "batch_size": batch_size,
+            "batch_sampler": None,
+            "num_workers": num_workers,
+            "collate_fn": collate_fn,
+            "pin_memory": pin_memory,
+            "timeout": 0,
+            "worker_init_fn": None,
+            "prefetch_factor": prefetch_factor,
+            "persistent_workers": persistent_workers,
+            "multiprocessing_context": "spawn" if num_workers > 0 else None,
+        }
 
     def setup(self, stage: str | None = None):
         """Called once to prepare the datasets."""

pvnet/models/base_model.py

@@ -23,6 +23,7 @@
     PYTORCH_WEIGHTS_NAME,
 )
 
+logger = logging.getLogger(__name__)
 
 def fill_config_paths_with_placeholder(config: dict, placeholder: str = "PLACEHOLDER") -> dict:
     """Modify the config in place to fill data paths with placeholder strings.
@@ -34,10 +35,9 @@ def fill_config_paths_with_placeholder(config: dict, placeholder: str = "PLACEHO
     input_config = config["input_data"]
 
     for source in ["generation", "satellite"]:
-        if source in input_config:
-            # If not empty - i.e. if used
-            if input_config[source]["zarr_path"] != "":
-                input_config[source]["zarr_path"] = f"{placeholder}.zarr"
+        # If source present but path not empty
+        if (source in input_config) and (input_config[source]["zarr_path"] != ""):
+            input_config[source]["zarr_path"] = f"{placeholder}.zarr"
 
     if "nwp" in input_config:
         for source in input_config["nwp"]:
@@ -101,8 +101,7 @@ def minimize_config_for_model(config: dict, model: "BaseModel") -> dict:
                 + (model.sat_encoder.sequence_length - 1) * sat_config["time_resolution_minutes"]
             )
 
-    if "pv" in input_config:
-        if not model.include_pv:
+    if ("pv" in input_config) and (not model.include_pv):
             del input_config["pv"]
 
     if "generation" in input_config:
@@ -166,11 +165,11 @@ def download_from_hf(
                 raise Exception(
                     f"Failed to download {filename} from {repo_id} after {max_retries} attempts."
                 ) from e
-            logging.warning(
-                (
+            logger.warning(
+
                     f"Attempt {attempt}/{max_retries} failed to download {filename} "
                     f"from {repo_id}. Retrying in {wait_time} seconds..."
-                )
+
             )
             time.sleep(wait_time)
 
@@ -343,7 +342,6 @@ def save_pretrained(
 
             print(message)
 
-        return
 
     @staticmethod
     def create_hugging_face_model_card(

pvnet/models/ensemble.py

@@ -42,7 +42,7 @@ def __init__(
             forecast_minutes,
             interval_minutes,
         ]:
-            assert all([p == param_list[0] for p in param_list]), param_list
+            assert all(p == param_list[0] for p in param_list), param_list
 
         super().__init__(
             history_minutes=history_minutes[0],

pvnet/models/late_fusion/encoders/basic_blocks.py

@@ -36,7 +36,6 @@ def __init__(
     @abstractmethod
     def forward(self):
         """Run model forward"""
-        pass
 
 
 class ResidualConv3dBlock(nn.Module):

pvnet/models/late_fusion/encoders/encoders3d.py

@@ -84,7 +84,7 @@ def __init__(
             ),
             nn.ELU(),
         ]
-        for _ in range(0, number_of_conv3d_layers - 1):
+        for _ in range(number_of_conv3d_layers - 1):
             conv_layers += [
                 nn.Conv3d(
                     in_channels=conv3d_channels,

pvnet/models/late_fusion/late_fusion.py

@@ -191,7 +191,7 @@ def __init__(
             if add_image_embedding_channel:
                 self.nwp_embed_dict = torch.nn.ModuleDict()
 
-            for nwp_source in nwp_encoders_dict.keys():
+            for nwp_source in nwp_encoders_dict:
                 nwp_sequence_len = (
                     nwp_history_minutes[nwp_source] // nwp_interval_minutes[nwp_source]
                     + nwp_forecast_minutes[nwp_source] // nwp_interval_minutes[nwp_source]

pvnet/models/late_fusion/linear_networks/basic_blocks.py

@@ -29,12 +29,11 @@ def cat_modes(self, x: OrderedDict[str, torch.Tensor] | torch.Tensor) -> torch.T
         elif isinstance(x, torch.Tensor):
             return x
         else:
-            raise ValueError(f"Input of unexpected type {type(x)}")
+            raise TypeError(f"Input of unexpected type {type(x)}")
 
     @abstractmethod
     def forward(self, x: OrderedDict[str, torch.Tensor] | torch.Tensor) -> torch.Tensor:
         """Run model forward"""
-        pass
 
 
 class ResidualLinearBlock(nn.Module):

pvnet/models/late_fusion/site_encoders/basic_blocks.py

@@ -33,4 +33,3 @@ def __init__(
     @abstractmethod
     def forward(self) -> torch.Tensor:
         """Run model forward"""
-        pass

pvnet/optimizers.py

@@ -67,7 +67,6 @@ class AbstractOptimizer(ABC):
     @abstractmethod
     def __call__(self, model: Module):
         """Abstract call"""
-        pass
 
 
 class Adam(AbstractOptimizer):
@@ -166,7 +165,7 @@ def _call_multi(self, model):
 
         group_args = []
 
-        for key in self.lr.keys():
+        for key in self.lr:
             if key == "default":
                 continue
 

pvnet/training/lightning_module.py

@@ -168,15 +168,15 @@ def _store_val_predictions(self, batch: TensorBatch, y_hat: torch.Tensor) -> Non
             y_hat = y_hat[..., None]
 
         ds_preds_batch = xr.Dataset(
-            data_vars=dict(
-                y_hat=(["sample_num", "forecast_step", "p_level"], y_hat),
-                y=(["sample_num", "forecast_step"], y),
-            ),
-            coords=dict(
-                ids=("sample_num", ids),
-                init_times_utc=("sample_num", init_times_utc),
-                p_level=p_levels,
-            ),
+            data_vars={
+                "y_hat": (["sample_num", "forecast_step", "p_level"], y_hat),
+                "y": (["sample_num", "forecast_step"], y),
+            },
+            coords={
+                "ids": ("sample_num", ids),
+                "init_times_utc": ("sample_num", init_times_utc),
+                "p_level": p_levels,
+            },
         )
         self.all_val_results.append(ds_preds_batch)
 

pvnet/training/train.py

@@ -64,13 +64,13 @@ def train(config: DictConfig) -> None:
     # Init lightning loggers
     loggers: list[Logger] = []
     if "logger" in config:
-        for _, lg_conf in config.logger.items():
+        for lg_conf in config.logger.values():
             loggers.append(hydra.utils.instantiate(lg_conf))
 
     # Init lightning callbacks
     callbacks: list[Callback] = []
     if "callbacks" in config:
-        for _, cb_conf in config.callbacks.items():
+        for cb_conf in config.callbacks.values():
             callbacks.append(hydra.utils.instantiate(cb_conf))
 
     # Align the wandb id with the checkpoint path

pyproject.toml

@@ -80,7 +80,10 @@ target-version = "py310"
 
 [tool.ruff.lint]
 extend-select = ["E", "D", "I"]
-ignore = ["D200","D202","D210","D212","D415","D105"]
+ignore = ["D200","D202","D210","D212","D415","D105",
+    "TRY002", # Create your own exception
+    "B008", # Do not perform function call typer.Argument
+    ]
 
 [tool.ruff.lint.mccabe]
 # Unlike Flake8, default to a complexity level of 10.

scripts/backtest.py

@@ -90,7 +90,7 @@ def populate_config_with_data_filepaths(config: dict, data_paths: dict) -> dict:
     # NWP is nested so must be treated separately
     if "nwp" in config["input_data"]:
         nwp_config = config["input_data"]["nwp"]
-        for nwp_source in nwp_config.keys():
+        for nwp_source in nwp_config:
             provider = nwp_config[nwp_source]["provider"]
             assert provider in data_paths["nwp"], f"Missing NWP path: {provider}"
             nwp_config[nwp_source]["zarr_path"] = data_paths["nwp"][provider]
@@ -325,11 +325,11 @@ def to_dataarray(
         """Put numpy array of predictions into a dataarray"""
 
         dims = ["init_time_utc", "location_id", "step"]
-        coords = dict(
-            init_time_utc=[t0],
-            location_id=location_ids,
-            step=self.steps,
-        )
+        coords = {
+            "init_time_utc": [t0],
+            "location_id": location_ids,
+            "step": self.steps,
+        }
 
         if output_quantiles is not None:
             dims.append("quantile")

scripts/migrate_old_model.py

@@ -73,14 +73,17 @@
 
 # This parameter has been removed
 if "target_key" in model_config:
-    if model_config["target_key"] == "site":
-        if "include_site_yield_history" in model_config:
+    if (
+        (model_config["target_key"] == "site") 
+        and ("include_site_yield_history" in model_config)
+        ):
             model_config["include_generation_history"] = model_config.pop(
                 "include_site_yield_history"
             )
 
-    if model_config["target_key"] == "gsp" or not model_config["target_key"]:
-        if "include_site_yield_history" in model_config:
+    if (
+        (model_config["target_key"] == "gsp" or not model_config["target_key"]) 
+        and ("include_site_yield_history" in model_config)):
             del model_config["include_site_yield_history"]
 
     del model_config["target_key"]
@@ -100,7 +103,7 @@
 
 # Re-find the model components in the new package structure
 if model_config.get("nwp_encoders_dict", None) is not None:
-    for k, v in model_config["nwp_encoders_dict"].items():
+    for v in model_config["nwp_encoders_dict"].values():
         v["_target_"] = (
             v["_target_"]
             .replace("multimodal", "late_fusion")
@@ -154,7 +157,7 @@
 
 # Add a note to the model card to say the model has been migrated
 with open(f"{save_dir}/{MODEL_CARD_NAME}", "a") as f:
-    current_date = datetime.date.today().strftime("%Y-%m-%d")
+    current_date = datetime.datetime.now(datetime.timezone.utc).date().strftime("%Y-%m-%d")
     pvnet_version = version("pvnet")
     f.write(
         f"\n\n---\n**Migration Note**: This model was migrated on {current_date} "

scripts/scorecard/generate_scorecard.py

@@ -136,14 +136,14 @@ def prep_ds(ds, y_true):
     backtest_ds_list = []
 
     # Add model name dimension to each backtest
-    for model in backtest_dict.keys():
-        if type(backtest_dict[model]) is list:
-            ds = xr.open_mfdataset(backtest_dict[model], engine="zarr")
+    for model, backtest_filepath in backtest_dict.items():
+        if type(backtest_filepath) is list:
+            ds = xr.open_mfdataset(backtest_filepath, engine="zarr")
             ds = ds.drop_duplicates(dim='init_time_utc')
         else:
-            ds = xr.open_zarr(backtest_dict[model])
+            ds = xr.open_zarr(backtest_filepath)
 
-        ds = ds.expand_dims(dict(model=[model]))
+        ds = ds.expand_dims({"model": [model]})
         backtest_ds_list.append(ds)
 
     # Intersect backtests by all dimensions except model name