PVNet 3.0.29__tar.gz → 3.0.35__tar.gz

{pvnet-3.0.29 → pvnet-3.0.35}/PKG-INFO

@@ -1,10 +1,8 @@
 Metadata-Version: 2.1
 Name: PVNet
-Version: 3.0.29
+Version: 3.0.35
 Summary: PVNet
-Author: Peter Dudfield
-Author-email: info@openclimatefix.org
-License: MIT
+Author-email: Peter Dudfield <info@openclimatefix.org>
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: ocf_datapipes>=3.3.6
@@ -33,9 +31,20 @@ Requires-Dist: python-dotenv
 Requires-Dist: hydra-optuna-sweeper
 Requires-Dist: rich
 Requires-Dist: gcsfs
+Provides-Extra: dev
+Requires-Dist: pvlive-api; extra == "dev"
+Requires-Dist: black; extra == "dev"
+Requires-Dist: flake8; extra == "dev"
+Requires-Dist: isort; extra == "dev"
+Requires-Dist: mypy; extra == "dev"
+Requires-Dist: pre-commit; extra == "dev"
+Requires-Dist: pytest; extra == "dev"
+Requires-Dist: pytest-cov; extra == "dev"
 Provides-Extra: all-models
 Requires-Dist: pytorch-tabnet; extra == "all-models"
 Requires-Dist: efficientnet_pytorch; extra == "all-models"
+Provides-Extra: all
+Requires-Dist: PVNet[all_models,dev]; extra == "all"
 
 # PVNet 2.1
 
@@ -70,7 +79,7 @@ Some slightly more structured notes on deliberate experiments we have performed
 ```bash
 git clone https://github.com/openclimatefix/PVNet.git
 cd PVNet
-pip install -r requirements.txt
+pip install .
 ```
 
 The commit history is extensive. To save download time, use a depth of 1:
@@ -88,7 +97,7 @@ pip install -e .
 ### Additional development dependencies
 
 ```bash
-pip install -r requirements-dev.txt
+pip install ".[dev]"
 ```
 
 
@@ -140,7 +149,7 @@ conda create -n ocf_datapipes python=3.10
 Then go inside the ocf_datapipes repo to add packages
 
 ```bash
-pip install -r requirements.txt -r requirements-dev.txt
+pip install ".[dev]"
 ```
 
 Then exit this environment, and enter back into the pvnet conda environment and install ocf_datapies in editable mode (-e). This means the package is directly linked to the source code in the ocf_datapies repo.
@@ -184,7 +193,7 @@ For this example we will be using the `streamed_batches.yaml` config. Like befor
 
 This should be given the whole path to the config on your local machine, such as for our example it should be changed to:
 
-`configuration: "/FULL-PATH-TO-REPO/PVNet/configs/datamodule/configuration/gcp_configuration.yaml"
+`configuration: "/FULL-PATH-TO-REPO/PVNet/configs/datamodule/configuration/gcp_configuration.yaml"`
 `
 
 Where `FULL-PATH-TO-REPO` represent the whole path to the PVNet repo on your local machine.
@@ -209,14 +218,14 @@ gcloud auth login
 
 For files stored in multiple locations they can be added as list. For example from the gcp_configuration.yaml file we can change from satellite data stored on a bucket:
 
-```
+```yaml
 satellite:
     satellite_zarr_path: gs://solar-pv-nowcasting-data/satellite/EUMETSAT/SEVIRI_RSS/v4/2020_nonhrv.zarr
 ```
 
 To satellite data hosted by Google:
 
-```
+```yaml
 satellite:
     satellite_zarr_paths:
       - "gs://public-datasets-eumetsat-solar-forecasting/satellite/EUMETSAT/SEVIRI_RSS/v4/2020_nonhrv.zarr"

{pvnet-3.0.29 → pvnet-3.0.35}/PVNet.egg-info/PKG-INFO

@@ -1,10 +1,8 @@
 Metadata-Version: 2.1
 Name: PVNet
-Version: 3.0.29
+Version: 3.0.35
 Summary: PVNet
-Author: Peter Dudfield
-Author-email: info@openclimatefix.org
-License: MIT
+Author-email: Peter Dudfield <info@openclimatefix.org>
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: ocf_datapipes>=3.3.6
@@ -33,9 +31,20 @@ Requires-Dist: python-dotenv
 Requires-Dist: hydra-optuna-sweeper
 Requires-Dist: rich
 Requires-Dist: gcsfs
+Provides-Extra: dev
+Requires-Dist: pvlive-api; extra == "dev"
+Requires-Dist: black; extra == "dev"
+Requires-Dist: flake8; extra == "dev"
+Requires-Dist: isort; extra == "dev"
+Requires-Dist: mypy; extra == "dev"
+Requires-Dist: pre-commit; extra == "dev"
+Requires-Dist: pytest; extra == "dev"
+Requires-Dist: pytest-cov; extra == "dev"
 Provides-Extra: all-models
 Requires-Dist: pytorch-tabnet; extra == "all-models"
 Requires-Dist: efficientnet_pytorch; extra == "all-models"
+Provides-Extra: all
+Requires-Dist: PVNet[all_models,dev]; extra == "all"
 
 # PVNet 2.1
 
@@ -70,7 +79,7 @@ Some slightly more structured notes on deliberate experiments we have performed
 ```bash
 git clone https://github.com/openclimatefix/PVNet.git
 cd PVNet
-pip install -r requirements.txt
+pip install .
 ```
 
 The commit history is extensive. To save download time, use a depth of 1:
@@ -88,7 +97,7 @@ pip install -e .
 ### Additional development dependencies
 
 ```bash
-pip install -r requirements-dev.txt
+pip install ".[dev]"
 ```
 
 
@@ -140,7 +149,7 @@ conda create -n ocf_datapipes python=3.10
 Then go inside the ocf_datapipes repo to add packages
 
 ```bash
-pip install -r requirements.txt -r requirements-dev.txt
+pip install ".[dev]"
 ```
 
 Then exit this environment, and enter back into the pvnet conda environment and install ocf_datapies in editable mode (-e). This means the package is directly linked to the source code in the ocf_datapies repo.
@@ -184,7 +193,7 @@ For this example we will be using the `streamed_batches.yaml` config. Like befor
 
 This should be given the whole path to the config on your local machine, such as for our example it should be changed to:
 
-`configuration: "/FULL-PATH-TO-REPO/PVNet/configs/datamodule/configuration/gcp_configuration.yaml"
+`configuration: "/FULL-PATH-TO-REPO/PVNet/configs/datamodule/configuration/gcp_configuration.yaml"`
 `
 
 Where `FULL-PATH-TO-REPO` represent the whole path to the PVNet repo on your local machine.
@@ -209,14 +218,14 @@ gcloud auth login
 
 For files stored in multiple locations they can be added as list. For example from the gcp_configuration.yaml file we can change from satellite data stored on a bucket:
 
-```
+```yaml
 satellite:
     satellite_zarr_path: gs://solar-pv-nowcasting-data/satellite/EUMETSAT/SEVIRI_RSS/v4/2020_nonhrv.zarr
 ```
 
 To satellite data hosted by Google:
 
-```
+```yaml
 satellite:
     satellite_zarr_paths:
       - "gs://public-datasets-eumetsat-solar-forecasting/satellite/EUMETSAT/SEVIRI_RSS/v4/2020_nonhrv.zarr"

{pvnet-3.0.29 → pvnet-3.0.35}/PVNet.egg-info/SOURCES.txt

@@ -2,9 +2,6 @@ LICENSE
 MANIFEST.in
 README.md
 pyproject.toml
-requirements-dev.txt
-requirements.txt
-setup.py
 PVNet.egg-info/PKG-INFO
 PVNet.egg-info/SOURCES.txt
 PVNet.egg-info/dependency_links.txt
@@ -12,6 +9,7 @@ PVNet.egg-info/requires.txt
 PVNet.egg-info/top_level.txt
 pvnet/__init__.py
 pvnet/callbacks.py
+pvnet/load_model.py
 pvnet/optimizers.py
 pvnet/training.py
 pvnet/utils.py
@@ -44,6 +42,4 @@ pvnet/models/multimodal/linear_networks/networks.py
 pvnet/models/multimodal/site_encoders/__init__.py
 pvnet/models/multimodal/site_encoders/basic_blocks.py
 pvnet/models/multimodal/site_encoders/encoders.py
-tests/__init__.py
-tests/conftest.py
 tests/test_end2end.py

{pvnet-3.0.29 → pvnet-3.0.35}/PVNet.egg-info/requires.txt

@@ -25,6 +25,19 @@ hydra-optuna-sweeper
 rich
 gcsfs
 
+[all]
+PVNet[all_models,dev]
+
 [all_models]
 pytorch-tabnet
 efficientnet_pytorch
+
+[dev]
+pvlive-api
+black
+flake8
+isort
+mypy
+pre-commit
+pytest
+pytest-cov

{pvnet-3.0.29 → pvnet-3.0.35}/PVNet.egg-info/top_level.txt

@@ -1,2 +1 @@
 pvnet
-tests

{pvnet-3.0.29 → pvnet-3.0.35}/README.md

@@ -31,7 +31,7 @@ Some slightly more structured notes on deliberate experiments we have performed
 ```bash
 git clone https://github.com/openclimatefix/PVNet.git
 cd PVNet
-pip install -r requirements.txt
+pip install .
 ```
 
 The commit history is extensive. To save download time, use a depth of 1:
@@ -49,7 +49,7 @@ pip install -e .
 ### Additional development dependencies
 
 ```bash
-pip install -r requirements-dev.txt
+pip install ".[dev]"
 ```
 
 
@@ -101,7 +101,7 @@ conda create -n ocf_datapipes python=3.10
 Then go inside the ocf_datapipes repo to add packages
 
 ```bash
-pip install -r requirements.txt -r requirements-dev.txt
+pip install ".[dev]"
 ```
 
 Then exit this environment, and enter back into the pvnet conda environment and install ocf_datapies in editable mode (-e). This means the package is directly linked to the source code in the ocf_datapies repo.
@@ -145,7 +145,7 @@ For this example we will be using the `streamed_batches.yaml` config. Like befor
 
 This should be given the whole path to the config on your local machine, such as for our example it should be changed to:
 
-`configuration: "/FULL-PATH-TO-REPO/PVNet/configs/datamodule/configuration/gcp_configuration.yaml"
+`configuration: "/FULL-PATH-TO-REPO/PVNet/configs/datamodule/configuration/gcp_configuration.yaml"`
 `
 
 Where `FULL-PATH-TO-REPO` represent the whole path to the PVNet repo on your local machine.
@@ -170,14 +170,14 @@ gcloud auth login
 
 For files stored in multiple locations they can be added as list. For example from the gcp_configuration.yaml file we can change from satellite data stored on a bucket:
 
-```
+```yaml
 satellite:
     satellite_zarr_path: gs://solar-pv-nowcasting-data/satellite/EUMETSAT/SEVIRI_RSS/v4/2020_nonhrv.zarr
 ```
 
 To satellite data hosted by Google:
 
-```
+```yaml
 satellite:
     satellite_zarr_paths:
       - "gs://public-datasets-eumetsat-solar-forecasting/satellite/EUMETSAT/SEVIRI_RSS/v4/2020_nonhrv.zarr"

pvnet-3.0.35/pvnet/__init__.py

@@ -0,0 +1,2 @@
+"""PVNet"""
+__version__ = "3.0.35"

pvnet-3.0.35/pvnet/load_model.py

@@ -0,0 +1,70 @@
+""" Load a model from its checkpoint directory """
+import glob
+import os
+
+import hydra
+import torch
+from pyaml_env import parse_config
+
+from pvnet.models.ensemble import Ensemble
+from pvnet.models.multimodal.unimodal_teacher import Model as UMTModel
+
+
+def get_model_from_checkpoints(
+    checkpoint_dir_paths: list[str],
+    val_best: bool = True,
+):
+    """Load a model from its checkpoint directory"""
+    is_ensemble = len(checkpoint_dir_paths) > 1
+
+    model_configs = []
+    models = []
+    data_configs = []
+
+    for path in checkpoint_dir_paths:
+        # Load the model
+        model_config = parse_config(f"{path}/model_config.yaml")
+
+        model = hydra.utils.instantiate(model_config)
+
+        if val_best:
+            # Only one epoch (best) saved per model
+            files = glob.glob(f"{path}/epoch*.ckpt")
+            if len(files) != 1:
+                raise ValueError(
+                    f"Found {len(files)} checkpoints @ {path}/epoch*.ckpt. Expected one."
+                )
+            checkpoint = torch.load(files[0], map_location="cpu")
+        else:
+            checkpoint = torch.load(f"{path}/last.ckpt", map_location="cpu")
+
+        model.load_state_dict(state_dict=checkpoint["state_dict"])
+
+        if isinstance(model, UMTModel):
+            model, model_config = model.convert_to_multimodal_model(model_config)
+
+        # Check for data config
+        data_config = f"{path}/data_config.yaml"
+
+        if os.path.isfile(data_config):
+            data_configs.append(data_config)
+        else:
+            data_configs.append(None)
+
+        model_configs.append(model_config)
+        models.append(model)
+
+    if is_ensemble:
+        model_config = {
+            "_target_": "pvnet.models.ensemble.Ensemble",
+            "model_list": model_configs,
+        }
+        model = Ensemble(model_list=models)
+        data_config = data_configs[0]
+
+    else:
+        model_config = model_configs[0]
+        model = models[0]
+        data_config = data_configs[0]
+
+    return model, model_config, data_config

{pvnet-3.0.29 → pvnet-3.0.35}/pvnet/utils.py

@@ -1,6 +1,5 @@
 """Utils"""
 import logging
-import os
 import warnings
 from collections.abc import Sequence
 from typing import Optional
@@ -13,7 +12,6 @@ import pylab
 import rich.syntax
 import rich.tree
 import xarray as xr
-import yaml
 from lightning.pytorch.loggers import Logger
 from lightning.pytorch.utilities import rank_zero_only
 from ocf_datapipes.batch import BatchKey
@@ -21,26 +19,6 @@ from ocf_datapipes.utils import Location
 from ocf_datapipes.utils.geospatial import osgb_to_lon_lat
 from omegaconf import DictConfig, OmegaConf
 
-import pvnet
-
-
-def load_config(config_file):
-    """
-    Open yam configruation file, and get rid eof '_target_' line
-    """
-
-    # get full path of config file
-    path = os.path.dirname(pvnet.__file__)
-    config_file = f"{path}/../{config_file}"
-
-    with open(config_file) as cfg:
-        config = yaml.load(cfg, Loader=yaml.FullLoader)
-
-    if "_target_" in config.keys():
-        config.pop("_target_")  # This is only for Hydra
-
-    return config
-
 
 def get_logger(name=__name__, level=logging.INFO) -> logging.Logger:
     """Initializes multi-GPU-friendly python logger."""
@@ -236,11 +214,10 @@ def finish(
     """Makes sure everything closed properly."""
 
     # without this sweeps with wandb logger might crash!
-    for logger in loggers:
-        if isinstance(logger, pl.loggers.wandb.WandbLogger):
-            import wandb
+    if any([isinstance(logger, pl.loggers.wandb.WandbLogger) for logger in loggers]):
+        import wandb
 
-            wandb.finish()
+        wandb.finish()
 
 
 def plot_batch_forecasts(

pvnet-3.0.35/pyproject.toml

@@ -0,0 +1,88 @@
+[project]
+name="PVNet"
+description = "PVNet"
+authors = [{name="Peter Dudfield", email="info@openclimatefix.org"}]
+dynamic = ["version", "readme"]
+license={file="LICENCE"}
+
+dependencies = [
+    "ocf_datapipes>=3.3.6",
+    "ocf_ml_metrics>=0.0.11",
+    "numpy",
+    "pandas",
+    "matplotlib",
+    "xarray",
+    "ipykernel",
+    "h5netcdf",
+    "torch>=2.0.0",
+    "lightning>=2.0.1",
+    "torchvision",
+    "pytest",
+    "pytest-cov",
+    "typer",
+    "sqlalchemy",
+    "fsspec[s3]",
+    "wandb",
+    "huggingface-hub",
+    "tensorboard",
+    "tqdm",
+    "omegaconf",
+    "hydra-core",
+    "python-dotenv",
+    "hydra-optuna-sweeper",
+    "rich",
+    "gcsfs",# gcsfs is only needed when getting data from Google Cloud Storage
+]
+
+[tool.setuptools.dynamic]
+version = {attr = "pvnet.__version__"}
+readme = {file = "README.md", content-type = "text/markdown"}
+
+[tool.setuptools.package-dir]
+"pvnet" = "pvnet"
+
+[project.optional-dependencies]
+dev=[
+    "pvlive-api",
+    "black",
+    "flake8",
+    "isort",
+    "mypy",
+    "pre-commit",
+    "pytest",
+    "pytest-cov",
+]
+all_models=[
+    "pytorch-tabnet",
+    "efficientnet_pytorch"
+]
+all=["PVNet[dev,all_models]"]
+
+[tool.mypy]
+exclude = [
+    "^tests/",
+]
+disallow_untyped_defs = true
+disallow_any_unimported = true
+no_implicit_optional = true
+check_untyped_defs = true
+warn_return_any = true
+warn_unused_ignores = true
+show_error_codes = true
+warn_unreachable = true
+
+[[tool.mypy.overrides]]
+module = [
+]
+ignore_missing_imports = true
+
+[tool.pytest.ini_options]
+minversion = "6.0"
+addopts = "-ra -q"
+testpaths = [
+    "tests",
+]
+
+[tool.ruff]
+line-length = 100
+exclude = [".ipynb_checkpoints", "tests", "configs.example"]

pvnet-3.0.29/pvnet/__init__.py

@@ -1,2 +0,0 @@
-"""PVNet"""
-__version__ = "3.0.29"

pvnet-3.0.29/pyproject.toml

@@ -1,28 +0,0 @@
-[tool.mypy]
-exclude = [
-    "^tests/",
-]
-disallow_untyped_defs = true
-disallow_any_unimported = true
-no_implicit_optional = true
-check_untyped_defs = true
-warn_return_any = true
-warn_unused_ignores = true
-show_error_codes = true
-warn_unreachable = true
-
-[[tool.mypy.overrides]]
-module = [
-]
-ignore_missing_imports = true
-
-[tool.pytest.ini_options]
-minversion = "6.0"
-addopts = "-ra -q"
-testpaths = [
-    "tests",
-]
-
-[tool.ruff]
-line-length = 100
-exclude = [".ipynb_checkpoints", "tests", "configs.example"]

pvnet-3.0.29/requirements-dev.txt

@@ -1,8 +0,0 @@
-pvlive-api @ git+https://github.com/SheffieldSolar/PV_Live-API
-black
-flake8
-isort
-mypy
-pre-commit
-pytest
-pytest-cov

pvnet-3.0.29/requirements.txt

@@ -1,27 +0,0 @@
-ocf_datapipes>=3.3.6
-ocf_ml_metrics>=0.0.11
-numpy
-pandas
-matplotlib
-xarray
-ipykernel
-h5netcdf
-torch>=2.0.0
-lightning>=2.0.1
-torchvision
-pytest
-pytest-cov
-typer
-sqlalchemy
-fsspec[s3]
-wandb
-huggingface-hub
-tensorboard
-tqdm
-omegaconf
-hydra-core
-python-dotenv
-hydra-optuna-sweeper
-rich
-# gcsfs is only needed when getting data from Google Cloud Storage
-gcsfs

pvnet-3.0.29/setup.py

@@ -1,33 +0,0 @@
-""" Usual setup file for package """
-# read the contents of your README file
-from pathlib import Path
-
-from setuptools import find_packages, setup
-
-this_directory = Path(__file__).parent
-long_description = (this_directory / "README.md").read_text()
-install_requires = (this_directory / "requirements.txt").read_text().splitlines()
-extras_require = {"all_models": ["pytorch-tabnet", "efficientnet_pytorch"]}
-
-with open("pvnet/__init__.py") as f:
-    for line in f:
-        if line.startswith("__version__"):
-            _, _, version = line.replace("'", "").split()
-            version = version.replace('"', "")
-
-
-setup(
-    name="PVNet",
-    version=version,
-    license="MIT",
-    description="PVNet",
-    author="Peter Dudfield",
-    author_email="info@openclimatefix.org",
-    company="Open Climate Fix Ltd",
-    install_requires=install_requires,
-    extras_require=extras_require,
-    long_description=long_description,
-    long_description_content_type="text/markdown",
-    include_package_data=True,
-    packages=find_packages(),
-)

pvnet-3.0.29/tests/conftest.py

@@ -1,305 +0,0 @@
-import os
-import glob
-import tempfile
-
-import pytest
-import pandas as pd
-import numpy as np
-import xarray as xr
-import torch
-import hydra
-
-from ocf_datapipes.batch import BatchKey
-from datetime import timedelta
-
-import pvnet
-from pvnet.data.datamodule import DataModule
-from pvnet.data.wind_datamodule import WindDataModule
-
-import pvnet.models.multimodal.encoders.encoders3d
-import pvnet.models.multimodal.linear_networks.networks
-import pvnet.models.multimodal.site_encoders.encoders
-from pvnet.models.multimodal.multimodal import Model
-
-
-xr.set_options(keep_attrs=True)
-
-
-def time_before_present(dt: timedelta):
-    return pd.Timestamp.now(tz=None) - dt
-
-
-@pytest.fixture
-def nwp_data():
-    # Load dataset which only contains coordinates, but no data
-    ds = xr.open_zarr(
-        f"{os.path.dirname(os.path.abspath(__file__))}/test_data/sample_data/nwp_shell.zarr"
-    )
-
-    # Last init time was at least 2 hours ago and hour to 3-hour interval
-    t0_datetime_utc = time_before_present(timedelta(hours=2)).floor(timedelta(hours=3))
-    ds.init_time.values[:] = pd.date_range(
-        t0_datetime_utc - timedelta(hours=3 * (len(ds.init_time) - 1)),
-        t0_datetime_utc,
-        freq=timedelta(hours=3),
-    )
-
-    # This is important to avoid saving errors
-    for v in list(ds.coords.keys()):
-        if ds.coords[v].dtype == object:
-            ds[v].encoding.clear()
-
-    for v in list(ds.variables.keys()):
-        if ds[v].dtype == object:
-            ds[v].encoding.clear()
-
-    # Add data to dataset
-    ds["UKV"] = xr.DataArray(
-        np.zeros([len(ds[c]) for c in ds.coords]),
-        coords=ds.coords,
-    )
-
-    # Add stored attributes to DataArray
-    ds.UKV.attrs = ds.attrs["_data_attrs"]
-    del ds.attrs["_data_attrs"]
-
-    return ds
-
-
-@pytest.fixture()
-def sat_data():
-    # Load dataset which only contains coordinates, but no data
-    ds = xr.open_zarr(
-        f"{os.path.dirname(os.path.abspath(__file__))}/test_data/sample_data/non_hrv_shell.zarr"
-    )
-
-    # Change times so they lead up to present. Delayed by at most 1 hour
-    t0_datetime_utc = time_before_present(timedelta(minutes=0)).floor(timedelta(minutes=30))
-    t0_datetime_utc = t0_datetime_utc - timedelta(minutes=30)
-    ds.time.values[:] = pd.date_range(
-        t0_datetime_utc - timedelta(minutes=5 * (len(ds.time) - 1)),
-        t0_datetime_utc,
-        freq=timedelta(minutes=5),
-    )
-
-    # Add data to dataset
-    ds["data"] = xr.DataArray(
-        np.zeros([len(ds[c]) for c in ds.coords]),
-        coords=ds.coords,
-    )
-
-    # Add stored attributes to DataArray
-    ds.data.attrs = ds.attrs["_data_attrs"]
-    del ds.attrs["_data_attrs"]
-
-    return ds
-
-
-@pytest.fixture()
-def sample_train_val_datamodule():
-    # duplicate the sample batcnes for more training/val data
-    n_duplicates = 10
-
-    with tempfile.TemporaryDirectory() as tmpdirname:
-        os.makedirs(f"{tmpdirname}/train")
-        os.makedirs(f"{tmpdirname}/val")
-
-        file_n = 0
-
-        for file in glob.glob("tests/test_data/sample_batches/train/*.pt"):
-            batch = torch.load(file)
-
-            for i in range(n_duplicates):
-                # Save fopr both train and val
-                torch.save(batch, f"{tmpdirname}/train/{file_n:06}.pt")
-                torch.save(batch, f"{tmpdirname}/val/{file_n:06}.pt")
-
-                file_n += 1
-
-        dm = DataModule(
-            configuration=None,
-            batch_size=2,
-            num_workers=0,
-            prefetch_factor=None,
-            train_period=[None, None],
-            val_period=[None, None],
-            test_period=[None, None],
-            batch_dir=f"{tmpdirname}",
-        )
-        yield dm
-
-
-@pytest.fixture()
-def sample_datamodule():
-    dm = DataModule(
-        configuration=None,
-        batch_size=2,
-        num_workers=0,
-        prefetch_factor=None,
-        train_period=[None, None],
-        val_period=[None, None],
-        test_period=[None, None],
-        batch_dir="tests/test_data/sample_batches",
-    )
-    return dm
-
-
-@pytest.fixture()
-def sample_batch(sample_datamodule):
-    batch = next(iter(sample_datamodule.train_dataloader()))
-    return batch
-
-
-@pytest.fixture()
-def sample_satellite_batch(sample_batch):
-    sat_image = sample_batch[BatchKey.satellite_actual]
-    return torch.swapaxes(sat_image, 1, 2)
-
-
-@pytest.fixture()
-def sample_pv_batch(sample_batch):
-    pv_data = sample_batch[BatchKey.pv]
-    return pv_data
-
-
-@pytest.fixture()
-def sample_wind_batch():
-    dm = WindDataModule(
-        configuration=None,
-        batch_size=2,
-        num_workers=0,
-        prefetch_factor=None,
-        train_period=[None, None],
-        val_period=[None, None],
-        test_period=[None, None],
-        batch_dir="tests/test_data/sample_wind_batches",
-    )
-    batch = next(iter(dm.train_dataloader()))
-    return batch
-
-
-@pytest.fixture()
-def model_minutes_kwargs():
-    kwargs = dict(
-        forecast_minutes=480,
-        history_minutes=120,
-    )
-    return kwargs
-
-
-@pytest.fixture()
-def encoder_model_kwargs():
-    # Used to test encoder model on satellite data
-    kwargs = dict(
-        sequence_length=(90 - 30) // 5 + 1,
-        image_size_pixels=24,
-        in_channels=11,
-        out_features=128,
-    )
-    return kwargs
-
-
-@pytest.fixture()
-def site_encoder_model_kwargs():
-    # Used to test site encoder model on PV data
-    kwargs = dict(
-        sequence_length=180 // 5 + 1,
-        num_sites=349,
-        out_features=128,
-    )
-    return kwargs
-
-
-@pytest.fixture()
-def site_encoder_sensor_model_kwargs():
-    # Used to test site encoder model on PV data
-    kwargs = dict(
-        sequence_length=180 // 5 + 1,
-        num_sites=26,
-        out_features=128,
-        num_channels=23,
-        target_key_to_use="wind",
-        input_key_to_use="sensor",
-    )
-    return kwargs
-
-
-@pytest.fixture()
-def raw_multimodal_model_kwargs(model_minutes_kwargs):
-    kwargs = dict(
-        sat_encoder=dict(
-            _target_=pvnet.models.multimodal.encoders.encoders3d.DefaultPVNet,
-            _partial_=True,
-            in_channels=11,
-            out_features=128,
-            number_of_conv3d_layers=6,
-            conv3d_channels=32,
-            image_size_pixels=24,
-        ),
-        nwp_encoders_dict={
-            "ukv": dict(
-                _target_=pvnet.models.multimodal.encoders.encoders3d.DefaultPVNet,
-                _partial_=True,
-                in_channels=2,
-                out_features=128,
-                number_of_conv3d_layers=6,
-                conv3d_channels=32,
-                image_size_pixels=24,
-            ),
-        },
-        add_image_embedding_channel=True,
-        pv_encoder=dict(
-            _target_=pvnet.models.multimodal.site_encoders.encoders.SingleAttentionNetwork,
-            _partial_=True,
-            num_sites=349,
-            out_features=40,
-            num_heads=4,
-            kdim=40,
-            id_embed_dim=20,
-        ),
-        output_network=dict(
-            _target_=pvnet.models.multimodal.linear_networks.networks.ResFCNet2,
-            _partial_=True,
-            fc_hidden_features=128,
-            n_res_blocks=6,
-            res_block_layers=2,
-            dropout_frac=0.0,
-        ),
-        embedding_dim=16,
-        include_sun=True,
-        include_gsp_yield_history=True,
-        sat_history_minutes=90,
-        nwp_history_minutes={"ukv": 120},
-        nwp_forecast_minutes={"ukv": 480},
-        pv_history_minutes=180,
-        min_sat_delay_minutes=30,
-    )
-
-    kwargs.update(model_minutes_kwargs)
-
-    return kwargs
-
-
-@pytest.fixture()
-def multimodal_model_kwargs(raw_multimodal_model_kwargs):
-    return hydra.utils.instantiate(raw_multimodal_model_kwargs)
-
-
-@pytest.fixture()
-def multimodal_model(multimodal_model_kwargs):
-    model = Model(**multimodal_model_kwargs)
-    return model
-
-
-@pytest.fixture()
-def multimodal_quantile_model(multimodal_model_kwargs):
-    model = Model(output_quantiles=[0.1, 0.5, 0.9], **multimodal_model_kwargs)
-    return model
-
-
-@pytest.fixture()
-def multimodal_weighted_quantile_model(multimodal_model_kwargs):
-    model = Model(
-        output_quantiles=[0.1, 0.5, 0.9], **multimodal_model_kwargs, use_weighted_loss=True
-    )
-    return model