ocf-data-sampler 0.0.24__py3-none-any.whl → 0.0.26__py3-none-any.whl

ocf_data_sampler/torch_datasets/valid_time_periods.py

@@ -0,0 +1,108 @@
+import numpy as np
+import pandas as pd
+
+from ocf_data_sampler.config import Configuration
+from ocf_data_sampler.select.find_contiguous_time_periods import find_contiguous_t0_periods_nwp, \
+    find_contiguous_t0_periods, intersection_of_multiple_dataframes_of_periods
+from ocf_data_sampler.time_functions import minutes
+
+
+def find_valid_time_periods(
+    datasets_dict: dict,
+    config: Configuration,
+):
+    """Find the t0 times where all of the requested input data is available
+
+    Args:
+        datasets_dict: A dictionary of input datasets
+        config: Configuration file
+    """
+
+    assert set(datasets_dict.keys()).issubset({"nwp", "sat", "gsp"})
+
+    contiguous_time_periods: dict[str: pd.DataFrame] = {}  # Used to store contiguous time periods from each data source
+
+    if "nwp" in datasets_dict:
+        for nwp_key, nwp_config in config.input_data.nwp.items():
+
+            da = datasets_dict["nwp"][nwp_key]
+
+            if nwp_config.dropout_timedeltas_minutes is None:
+                max_dropout = minutes(0)
+            else:
+                max_dropout = minutes(np.max(np.abs(nwp_config.dropout_timedeltas_minutes)))
+
+            if nwp_config.max_staleness_minutes is None:
+                max_staleness = None
+            else:
+                max_staleness = minutes(nwp_config.max_staleness_minutes)
+
+            # The last step of the forecast is lost if we have to diff channels
+            if len(nwp_config.accum_channels) > 0:
+                end_buffer = minutes(nwp_config.time_resolution_minutes)
+            else:
+                end_buffer = minutes(0)
+
+            # This is the max staleness we can use considering the max step of the input data
+            max_possible_staleness = (
+                pd.Timedelta(da["step"].max().item())
+                - minutes(nwp_config.forecast_minutes)
+                - end_buffer
+            )
+
+            # Default to use max possible staleness unless specified in config
+            if max_staleness is None:
+                max_staleness = max_possible_staleness
+            else:
+                # Make sure the max acceptable staleness isn't longer than the max possible
+                assert max_staleness <= max_possible_staleness
+
+            time_periods = find_contiguous_t0_periods_nwp(
+                datetimes=pd.DatetimeIndex(da["init_time_utc"]),
+                history_duration=minutes(nwp_config.history_minutes),
+                max_staleness=max_staleness,
+                max_dropout=max_dropout,
+            )
+
+            contiguous_time_periods[f'nwp_{nwp_key}'] = time_periods
+
+    if "sat" in datasets_dict:
+        sat_config = config.input_data.satellite
+
+        time_periods = find_contiguous_t0_periods(
+            pd.DatetimeIndex(datasets_dict["sat"]["time_utc"]),
+            sample_period_duration=minutes(sat_config.time_resolution_minutes),
+            history_duration=minutes(sat_config.history_minutes),
+            forecast_duration=minutes(sat_config.forecast_minutes),
+        )
+
+        contiguous_time_periods['sat'] = time_periods
+
+    if "gsp" in datasets_dict:
+        gsp_config = config.input_data.gsp
+
+        time_periods = find_contiguous_t0_periods(
+            pd.DatetimeIndex(datasets_dict["gsp"]["time_utc"]),
+            sample_period_duration=minutes(gsp_config.time_resolution_minutes),
+            history_duration=minutes(gsp_config.history_minutes),
+            forecast_duration=minutes(gsp_config.forecast_minutes),
+        )
+
+        contiguous_time_periods['gsp'] = time_periods
+
+    # just get the values (not the keys)
+    contiguous_time_periods_values = list(contiguous_time_periods.values())
+
+    # Find joint overlapping contiguous time periods
+    if len(contiguous_time_periods_values) > 1:
+        valid_time_periods = intersection_of_multiple_dataframes_of_periods(
+            contiguous_time_periods_values
+        )
+    else:
+        valid_time_periods = contiguous_time_periods_values[0]
+
+    # check there are some valid time periods
+    if len(valid_time_periods) == 0:
+        raise ValueError(f"No valid time periods found, {contiguous_time_periods=}")
+
+    return valid_time_periods

{ocf_data_sampler-0.0.24.dist-info → ocf_data_sampler-0.0.26.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: ocf_data_sampler
-Version: 0.0.24
+Version: 0.0.26
 Summary: Sample from weather data for renewable energy prediction
 Author: James Fulton, Peter Dudfield, and the Open Climate Fix team
 Author-email: info@openclimatefix.org

{ocf_data_sampler-0.0.24.dist-info → ocf_data_sampler-0.0.26.dist-info}/RECORD

@@ -1,41 +1,52 @@
 ocf_data_sampler/__init__.py,sha256=AbpHGcgLb-kRsJGnwFEktk7uzpZOCcBY74-YBdrKVGs,1
 ocf_data_sampler/constants.py,sha256=tUwHrsGShqIn5Izze4i32_xB6X0v67rvQwIYB-P5PJQ,3355
+ocf_data_sampler/time_functions.py,sha256=R6ZlVEe6h4UlJeUW7paZYAMWveOv9MTjMsoISCwnsiE,284
 ocf_data_sampler/config/__init__.py,sha256=YXnAkgHViHB26hSsjiv32b6EbpG-A1kKTkARJf0_RkY,212
 ocf_data_sampler/config/load.py,sha256=4f7vPHAIAmd-55tPxoIzn7F_TI_ue4NxkDcLPoVWl0g,943
-ocf_data_sampler/config/model.py,sha256=bvU3BEMtcUh-N17fMVLTYtN-J2GcTM9Qq-CI5AfbE4Q,8128
+ocf_data_sampler/config/model.py,sha256=YnGOzt6T835h6bozWqrlMnUIHPo26U8o-DTKAKvv_24,7121
 ocf_data_sampler/config/save.py,sha256=wKdctbv0dxIIiQtcRHLRxpWQVhEFQ_FCWg-oNaRLIps,1093
 ocf_data_sampler/data/uk_gsp_locations.csv,sha256=RSh7DRh55E3n8lVAaWXGTaXXHevZZtI58td4d4DhGos,10415772
 ocf_data_sampler/load/__init__.py,sha256=MjgfxilTzyz1RYFoBEeAXmE9hyjknLvdmlHPmlAoiQY,44
 ocf_data_sampler/load/gsp.py,sha256=Gcr1JVUOPKhFRDCSHtfPDjxx0BtyyEhXrZvGEKLPJ5I,759
+ocf_data_sampler/load/load_dataset.py,sha256=Ua3RaUg4PIYJkD9BKqTfN8IWUbezbhThJGgEkd9PcaE,1587
 ocf_data_sampler/load/satellite.py,sha256=3KlA1fx4SwxdzM-jC1WRaONXO0D6m0WxORnEnwUnZrA,2967
+ocf_data_sampler/load/site.py,sha256=ROif2XXIIgBz-JOOiHymTq1CMXswJ3AzENU9DJmYpcU,782
 ocf_data_sampler/load/utils.py,sha256=EQGvVWlGMoSOdbDYuMfVAa0v6wmAOPmHIAemdrTB5v4,1406
 ocf_data_sampler/load/nwp/__init__.py,sha256=SmcrnbygO5xtCKmGR4wtHrj-HI7nOAvnAtfuvRufBGQ,25
 ocf_data_sampler/load/nwp/nwp.py,sha256=O4QnajEZem8BvBgTcYYDBhRhgqPYuJkolHmpMRmrXEA,610
 ocf_data_sampler/load/nwp/providers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-ocf_data_sampler/load/nwp/providers/ecmwf.py,sha256=vW-p3vCyQ-CofKo555-gE7VDi5hlpjtjTLfHqWF0HEE,1175
+ocf_data_sampler/load/nwp/providers/ecmwf.py,sha256=2iR1Iy542lo51rC6XFLV-3pbUE68dWjlHa6TVJzx3ac,1280
 ocf_data_sampler/load/nwp/providers/ukv.py,sha256=79Bm7q-K_GJPYMy62SUIZbRWRF4-tIaB1dYPEgLD9vo,1207
 ocf_data_sampler/load/nwp/providers/utils.py,sha256=Sy2exG1wpXLLhMXYdsfR-DZMR3txG1_bBmBdchlc-yA,848
-ocf_data_sampler/numpy_batch/__init__.py,sha256=M9b7L7kSgoZt2FKENDe-uFI_Qzs-fDecw-5qyrhnTm4,290
+ocf_data_sampler/numpy_batch/__init__.py,sha256=8MgRF29rK9bKP4b4iHakaoGwBKUcjWZ-VFKjCcq53QA,336
 ocf_data_sampler/numpy_batch/gsp.py,sha256=QjQ25JmtufvdiSsxUkBTPhxouYGWPnnWze8pXr_aBno,960
 ocf_data_sampler/numpy_batch/nwp.py,sha256=dAehfRo5DL2Yb20ifHHl5cU1QOrm3ZOpQmN39fSUOw8,1255
 ocf_data_sampler/numpy_batch/satellite.py,sha256=3NoE_ElzMHwO60apqJeFAwI6J7eIxD0OWTyAVl-uJi8,903
+ocf_data_sampler/numpy_batch/site.py,sha256=lJYMEot50UgSBnSOgADQMjUhky1YyWKYqwNsisyYV6w,789
 ocf_data_sampler/numpy_batch/sun_position.py,sha256=zw2bjtcjsm_tvKk0r_MZmgfYUJLHuLjLly2sMjwP3XI,1606
-ocf_data_sampler/select/__init__.py,sha256=AbpHGcgLb-kRsJGnwFEktk7uzpZOCcBY74-YBdrKVGs,1
-ocf_data_sampler/select/dropout.py,sha256=zDpVLMjGb70RRyYKN-WI2Kp3x9SznstT4cMcZ4dsvJg,1066
+ocf_data_sampler/select/__init__.py,sha256=E4AJulEbO2K-o0UlG1fgaEteuf_1ZFjHTvrotXSb4YU,332
+ocf_data_sampler/select/dropout.py,sha256=HCx5Wzk8Oh2Z9vV94Jy-ALJsHtGduwvMaQOleQXp5z0,1142
 ocf_data_sampler/select/fill_time_periods.py,sha256=iTtMjIPFYG5xtUYYedAFBLjTWWUa7t7WQ0-yksWf0-E,440
 ocf_data_sampler/select/find_contiguous_time_periods.py,sha256=6ioB8LeFpFNBMgKDxrgG3zqzNjkBF_jlV9yye2ZYT2E,11925
-ocf_data_sampler/select/geospatial.py,sha256=oHJoKEKubn3v3yKCVeuiPxuGroVA4RyrpNi6ARq5woE,3558
+ocf_data_sampler/select/geospatial.py,sha256=4xL-9y674jjoaXeqE52NHCHVfknciE4OEGsZtn9DvP4,4911
 ocf_data_sampler/select/location.py,sha256=26Y5ZjfFngShBwXieuWSoOA-RLaRzci4TTmcDk3Wg7U,2015
-ocf_data_sampler/select/select_spatial_slice.py,sha256=hWIJe4_VzuQ2iiiQh7V17AXwTILT5kIkUvzG458J_Gw,11220
+ocf_data_sampler/select/select_spatial_slice.py,sha256=WNxwur9Q5oetvogATw8-hNejDuEwrXHzuZIovFDjNJA,11488
 ocf_data_sampler/select/select_time_slice.py,sha256=41cch1fQr59fZgv7UHsNGc3OvoynrixT3bmr3_1d7cU,6628
+ocf_data_sampler/select/spatial_slice_for_dataset.py,sha256=3tRrMBXr7s4CnClbVSIq7hpls3H4Y3qYTDwswcxCCCE,1763
+ocf_data_sampler/select/time_slice_for_dataset.py,sha256=5gcTGgQ1D524OhullNRWq3hxCwl2SoliGR210G-62JA,4216
 ocf_data_sampler/torch_datasets/__init__.py,sha256=AbpHGcgLb-kRsJGnwFEktk7uzpZOCcBY74-YBdrKVGs,1
-ocf_data_sampler/torch_datasets/pvnet_uk_regional.py,sha256=Mlz_uyt8c8-uN0uaEiJV2DgF5WAqtWlsINFgA925CZI,19025
+ocf_data_sampler/torch_datasets/process_and_combine.py,sha256=GA-tGZLEMNAqX5Zun_7tPcTWVxlVtwejC9zfXPECwSk,4989
+ocf_data_sampler/torch_datasets/pvnet_uk_regional.py,sha256=TpHALGU7hpo3iLbvD0nkoY6zu94Vq99W1V1qSGEcIW8,5552
+ocf_data_sampler/torch_datasets/site.py,sha256=1k0fWXYwAAIWG5DX_j3tgNfY8gglfPGLNzNlZd8EnJs,6631
+ocf_data_sampler/torch_datasets/valid_time_periods.py,sha256=vP25e7DpWAu4dACTFMJZm0bi304iUFdi1XySAmxi_c0,4159
+scripts/refactor_site.py,sha256=asZ27hQ4IyXgCCUaFJqcz1ObBNcV2W3ywqHBpSXA_fc,1728
 tests/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-tests/conftest.py,sha256=O77dmow8mGpGPbZ6Pz7ma7cLaiV1k8mxW1eYg37Avrw,5585
-tests/config/test_config.py,sha256=G_PD_pXib0zdRBPUIn0jjwJ9VyoKaO_TanLN1Mh5Ca4,5055
+tests/conftest.py,sha256=ZRktySCynj3NBbFRR4EFNLRLFMErkQsC-qQlmQzhbRg,7360
+tests/config/test_config.py,sha256=C8NppoEVCMKxTTUf3o_z1Jb_I2DDH75XKpQ9x45U3Hw,5090
 tests/load/test_load_gsp.py,sha256=aT_nqaSXmUTcdHzuTT7AmXJr3R31k4OEN-Fv3eLxlQE,424
 tests/load/test_load_nwp.py,sha256=3qyyDkB1q9t3tyAwogfotNrxqUOpXXimco1CImoEWGg,753
 tests/load/test_load_satellite.py,sha256=STX5AqqmOAgUgE9R1xyq_sM3P1b8NKdGjO-hDhayfxM,524
+tests/load/test_load_sites.py,sha256=T9lSEnGPI8FQISudVYHHNTHeplNS62Vrx48jaZ6J_Jo,364
 tests/numpy_batch/test_gsp.py,sha256=VANXV32K8aLX4dCdhCUnDorJmyNN-Bjc7Wc1N-RzWEk,548
 tests/numpy_batch/test_nwp.py,sha256=Fnj7cR-VR2Z0kMu8SrgnIayjxWnPWrYFjWSjMmnrh4Y,1445
 tests/numpy_batch/test_satellite.py,sha256=8a4ZwMLpsOmYKmwI1oW_su_hwkCNYMEJAEfa0dbsx1k,1179
@@ -46,9 +57,10 @@ tests/select/test_find_contiguous_time_periods.py,sha256=G6tJRJd0DMfH9EdfzlKWsmf
 tests/select/test_location.py,sha256=_WZk2FPYeJ-nIfCJS6Sp_yaVEEo7m31DmMFoZzgyCts,2712
 tests/select/test_select_spatial_slice.py,sha256=7EX9b6g-pMdACQx3yefjs5do2s-Rho2UmKevV4oglsU,5147
 tests/select/test_select_time_slice.py,sha256=XC1J3DBBDnt81jcba5u-Hnd0yKv8GIQErLm-OECV6rs,10147
-tests/torch_datasets/test_pvnet_uk_regional.py,sha256=u3taw6p3oozM0_7cEEhCYbImAQPRldRhpruqSyV08Vg,2675
-ocf_data_sampler-0.0.24.dist-info/LICENSE,sha256=F-Q3UFCR-BECSocV55BFDpn4YKxve9PKrm-lTt6o_Tg,1073
-ocf_data_sampler-0.0.24.dist-info/METADATA,sha256=wwIltvHOvOd-L2KaOF3jsLOMx-QuY6yP6sNR0QddCdk,5269
-ocf_data_sampler-0.0.24.dist-info/WHEEL,sha256=OVMc5UfuAQiSplgO0_WdW7vXVGAt9Hdd6qtN4HotdyA,91
-ocf_data_sampler-0.0.24.dist-info/top_level.txt,sha256=KaQn5qzkJGJP6hKWqsVAc9t0cMLjVvSTk8-kTrW79SA,23
-ocf_data_sampler-0.0.24.dist-info/RECORD,,
+tests/torch_datasets/test_pvnet_uk_regional.py,sha256=8gxjJO8FhY-ImX6eGnihDFsa8fhU2Zb4bVJaToJwuwo,2653
+tests/torch_datasets/test_site.py,sha256=yTv6tAT6lha5yLYJiC8DNms1dct8o_ObPV97dHZyT7I,2719
+ocf_data_sampler-0.0.26.dist-info/LICENSE,sha256=F-Q3UFCR-BECSocV55BFDpn4YKxve9PKrm-lTt6o_Tg,1073
+ocf_data_sampler-0.0.26.dist-info/METADATA,sha256=VRnSRX4dgDbz4k9bwSM66uqaHI4P97xC97_NsEIt5qU,5269
+ocf_data_sampler-0.0.26.dist-info/WHEEL,sha256=P9jw-gEje8ByB7_hXoICnHtVCrEwMQh-630tKvQWehc,91
+ocf_data_sampler-0.0.26.dist-info/top_level.txt,sha256=Faob6N6cFdPc5eUpCTYcXgCaNhi4XLLteUL5W5ayYmg,31
+ocf_data_sampler-0.0.26.dist-info/RECORD,,

{ocf_data_sampler-0.0.24.dist-info → ocf_data_sampler-0.0.26.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (75.2.0)
+Generator: setuptools (75.3.0)
 Root-Is-Purelib: true
 Tag: py3-none-any
 

{ocf_data_sampler-0.0.24.dist-info → ocf_data_sampler-0.0.26.dist-info}/top_level.txt

@@ -1,2 +1,3 @@
 ocf_data_sampler
+scripts
 tests

scripts/refactor_site.py

@@ -0,0 +1,50 @@
+""" Helper functions for refactoring legacy site data """
+
+
+def legacy_format(data_ds, metadata_df):
+    """This formats old legacy data to the new format.
+
+    1. This renames the columns in the metadata
+    2. Re-formats the site data from data variables named by the site_id to
+    a data array with a site_id dimension. Also adds capacity_kwp to the dataset as a time series for each site_id
+    """
+
+    if "system_id" in metadata_df.columns:
+        metadata_df["site_id"] = metadata_df["system_id"]
+
+    if "capacity_megawatts" in metadata_df.columns:
+        metadata_df["capacity_kwp"] = metadata_df["capacity_megawatts"] * 1000
+
+    # only site data has the site_id as data variables.
+    # We want to join them all together and create another coordinate called site_id
+    if "0" in data_ds:
+        gen_df = data_ds.to_dataframe()
+        gen_da = xr.DataArray(
+            data=gen_df.values,
+            coords=(
+                ("time_utc", gen_df.index.values),
+                ("site_id", metadata_df["site_id"]),
+            ),
+            name="generation_kw",
+        )
+
+        capacity_df = gen_df
+        for col in capacity_df.columns:
+            capacity_df[col] = metadata_df[metadata_df["site_id"].astype(str) == col][
+                "capacity_kwp"
+            ].iloc[0]
+        capacity_da = xr.DataArray(
+            data=capacity_df.values,
+            coords=(
+                ("time_utc", gen_df.index.values),
+                ("site_id", metadata_df["site_id"]),
+            ),
+            name="capacity_kwp",
+        )
+        data_ds = xr.Dataset(
+            {
+                "generation_kw": gen_da,
+                "capacity_kwp": capacity_da,
+            }
+        )
+    return data_ds

tests/config/test_config.py

@@ -10,13 +10,13 @@ from ocf_data_sampler.config import (
 )
 
 
-def test_default():
+def test_default_configuration():
     """Test default pydantic class"""
 
     _ = Configuration()
 
 
-def test_yaml_load_test_config(test_config_filename):
+def test_load_yaml_configuration(test_config_filename):
     """
     Test that yaml loading works for 'test_config.yaml'
     and fails for an empty .yaml file
@@ -56,7 +56,7 @@ def test_yaml_save(test_config_filename):
         assert test_config == tmp_config
 
 
-def test_extra_field():
+def test_extra_field_error():
     """
     Check an extra parameters in config causes error
     """
@@ -99,10 +99,11 @@ def test_incorrect_nwp_provider(test_config_filename):
 
     configuration = load_yaml_configuration(test_config_filename)
 
-    configuration.input_data.nwp['ukv'].nwp_provider = "unexpected_provider"
+    configuration.input_data.nwp['ukv'].provider = "unexpected_provider"
     with pytest.raises(Exception, match="NWP provider"):
         _ = Configuration(**configuration.model_dump())
 
+
 def test_incorrect_dropout(test_config_filename):
     """
     Check a dropout timedelta over 0 causes error and 0 doesn't
@@ -119,6 +120,7 @@ def test_incorrect_dropout(test_config_filename):
     configuration.input_data.nwp['ukv'].dropout_timedeltas_minutes = [0]
     _ = Configuration(**configuration.model_dump())
 
+
 def test_incorrect_dropout_fraction(test_config_filename):
     """
     Check dropout fraction outside of range causes error
@@ -127,11 +129,12 @@ def test_incorrect_dropout_fraction(test_config_filename):
     configuration = load_yaml_configuration(test_config_filename)
 
     configuration.input_data.nwp['ukv'].dropout_fraction= 1.1
-    with pytest.raises(Exception, match="Dropout fraction must be between 0 and 1"):
+
+    with pytest.raises(ValidationError,  match="Input should be less than or equal to 1"):
         _ = Configuration(**configuration.model_dump())
 
     configuration.input_data.nwp['ukv'].dropout_fraction= -0.1
-    with pytest.raises(Exception, match="Dropout fraction must be between 0 and 1"):
+    with pytest.raises(ValidationError, match="Input should be greater than or equal to 0"):
         _ = Configuration(**configuration.model_dump())
 
 

tests/conftest.py

@@ -6,6 +6,8 @@ import pytest
 import xarray as xr
 import tempfile
 
+from ocf_data_sampler.config.model import Site
+
 _top_test_directory = os.path.dirname(os.path.realpath(__file__))
 
 @pytest.fixture()
@@ -197,6 +199,66 @@ def ds_uk_gsp():
     })
 
 
+@pytest.fixture(scope="session")
+def data_sites() -> Site:
+    """
+    Make fake data for sites
+    Returns: filename for netcdf file, and csv metadata
+    """
+    times = pd.date_range("2023-01-01 00:00", "2023-01-02 00:00", freq="30min")
+    site_ids = list(range(0,10))
+    capacity_kwp_1d = np.array([0.1,1.1,4,6,8,9,15,2,3,4])
+    # these are quite specific for the fake satellite data
+    longitude = np.arange(-4, -3, 0.1)
+    latitude = np.arange(51, 52, 0.1)
+
+    generation = np.random.uniform(0, 200, size=(len(times), len(site_ids))).astype(np.float32)
+
+    # repeat capacity in new dims len(times) times
+    capacity_kwp = (np.tile(capacity_kwp_1d, len(times))).reshape(len(times),10)
+
+    coords = (
+        ("time_utc", times),
+        ("site_id", site_ids),
+    )
+
+    da_cap = xr.DataArray(
+        capacity_kwp,
+        coords=coords,
+    )
+
+    da_gen = xr.DataArray(
+        generation,
+        coords=coords,
+    )
+
+    # metadata
+    meta_df = pd.DataFrame(columns=[], data = [])
+    meta_df['site_id'] = site_ids
+    meta_df['capacity_kwp'] = capacity_kwp_1d
+    meta_df['longitude'] = longitude
+    meta_df['latitude'] = latitude
+
+    generation = xr.Dataset({
+        "capacity_kwp": da_cap,
+        "generation_kw": da_gen,
+    })
+
+    with tempfile.TemporaryDirectory() as tmpdir:
+        filename = tmpdir + "/sites.netcdf"
+        filename_csv = tmpdir + "/sites_metadata.csv"
+        generation.to_netcdf(filename)
+        meta_df.to_csv(filename_csv)
+
+        site = Site(file_path=filename,
+                    metadata_file_path=filename_csv,
+                    time_resolution_minutes=30,
+                    forecast_minutes=60,
+                    history_minutes=30)
+
+        yield site
+
+
 @pytest.fixture(scope="session")
 def uk_gsp_zarr_path(ds_uk_gsp):
 

tests/load/test_load_sites.py

@@ -0,0 +1,14 @@
+from ocf_data_sampler.load.site import open_site
+import xarray as xr
+
+
+def test_open_site(data_sites):
+    da = open_site(data_sites)
+
+    assert isinstance(da, xr.DataArray)
+    assert da.dims == ("time_utc", "site_id")
+
+    assert "capacity_kwp" in da.coords
+    assert "latitude" in da.coords
+    assert "longitude" in da.coords
+    assert da.shape == (49, 10)

tests/torch_datasets/test_pvnet_uk_regional.py

@@ -11,9 +11,9 @@ def pvnet_config_filename(tmp_path, config_filename, nwp_ukv_zarr_path, uk_gsp_z
 
     # adjust config to point to the zarr file
     config = load_yaml_configuration(config_filename)
-    config.input_data.nwp['ukv'].nwp_zarr_path = nwp_ukv_zarr_path
-    config.input_data.satellite.satellite_zarr_path = sat_zarr_path
-    config.input_data.gsp.gsp_zarr_path = uk_gsp_zarr_path
+    config.input_data.nwp['ukv'].zarr_path = nwp_ukv_zarr_path
+    config.input_data.satellite.zarr_path = sat_zarr_path
+    config.input_data.gsp.zarr_path = uk_gsp_zarr_path
 
     filename = f"{tmp_path}/configuration.yaml"
     save_yaml_configuration(config, filename)
@@ -60,7 +60,7 @@ def test_pvnet_no_gsp(pvnet_config_filename):
     # load config
     config = load_yaml_configuration(pvnet_config_filename)
     # remove gsp
-    config.input_data.gsp.gsp_zarr_path = ''
+    config.input_data.gsp.zarr_path = ''
 
     # save temp config file
     with tempfile.NamedTemporaryFile() as temp_config_file:

tests/torch_datasets/test_site.py

@@ -0,0 +1,85 @@
+import pandas as pd
+import pytest
+
+from ocf_data_sampler.torch_datasets.site import SitesDataset
+from ocf_data_sampler.config import load_yaml_configuration, save_yaml_configuration
+from ocf_data_sampler.numpy_batch.nwp import NWPBatchKey
+from ocf_data_sampler.numpy_batch.site import SiteBatchKey
+from ocf_data_sampler.numpy_batch.satellite import SatelliteBatchKey
+
+
+@pytest.fixture()
+def site_config_filename(tmp_path, config_filename, nwp_ukv_zarr_path, sat_zarr_path, data_sites):
+
+    # adjust config to point to the zarr file
+    config = load_yaml_configuration(config_filename)
+    config.input_data.nwp["ukv"].zarr_path = nwp_ukv_zarr_path
+    config.input_data.satellite.zarr_path = sat_zarr_path
+    config.input_data.site = data_sites
+    config.input_data.gsp = None
+
+    filename = f"{tmp_path}/configuration.yaml"
+    save_yaml_configuration(config, filename)
+    return filename
+
+
+def test_site(site_config_filename):
+
+    # Create dataset object
+    dataset = SitesDataset(site_config_filename)
+
+    assert len(dataset) == 10 * 41
+    # TODO check 41
+
+    # Generate a sample
+    sample = dataset[0]
+
+    assert isinstance(sample, dict)
+
+    for key in [
+        NWPBatchKey.nwp,
+        SatelliteBatchKey.satellite_actual,
+        SiteBatchKey.generation,
+        SiteBatchKey.site_solar_azimuth,
+        SiteBatchKey.site_solar_elevation,
+    ]:
+        assert key in sample
+
+    for nwp_source in ["ukv"]:
+        assert nwp_source in sample[NWPBatchKey.nwp]
+
+    # check the shape of the data is correct
+    # 30 minutes of 5 minute data (inclusive), one channel, 2x2 pixels
+    assert sample[SatelliteBatchKey.satellite_actual].shape == (7, 1, 2, 2)
+    # 3 hours of 60 minute data (inclusive), one channel, 2x2 pixels
+    assert sample[NWPBatchKey.nwp]["ukv"][NWPBatchKey.nwp].shape == (4, 1, 2, 2)
+    # 3 hours of 30 minute data (inclusive)
+    assert sample[SiteBatchKey.generation].shape == (4,)
+    # Solar angles have same shape as GSP data
+    assert sample[SiteBatchKey.site_solar_azimuth].shape == (4,)
+    assert sample[SiteBatchKey.site_solar_elevation].shape == (4,)
+
+
+def test_site_time_filter_start(site_config_filename):
+
+    # Create dataset object
+    dataset = SitesDataset(site_config_filename, start_time="2024-01-01")
+
+    assert len(dataset) == 0
+
+
+def test_site_time_filter_end(site_config_filename):
+
+    # Create dataset object
+    dataset = SitesDataset(site_config_filename, end_time="2000-01-01")
+
+    assert len(dataset) == 0
+
+
+def test_site_get_sample(site_config_filename):
+
+    # Create dataset object
+    dataset = SitesDataset(site_config_filename)
+
+    assert len(dataset) == 410
+    sample = dataset.get_sample(t0=pd.Timestamp("2023-01-01 12:00"), site_id=1)