Handling Leap Years in OpenMeteo Resource Data

h2integrate/resource/resource_base.py

@@ -90,6 +90,7 @@ def setup(self):
         self.resource_data = None
         self.resource_site = [self.config.latitude, self.config.longitude]
         self.dt = self.options["plant_config"]["plant"]["simulation"]["dt"]
+        self.n_timesteps = self.options["plant_config"]["plant"]["simulation"]["n_timesteps"]
         self.add_input("latitude", self.config.latitude, units="deg")
         self.add_input("longitude", self.config.longitude, units="deg")
 

h2integrate/resource/solar/openmeteo_solar.py

@@ -21,6 +21,8 @@ class OpenMeteoHistoricalSolarAPIConfig(ResourceBaseAPIConfig):
     Args:
         resource_year (int): Year to use for resource data.
             Must been between 1940 the year before the current calendar year. (inclusive).
+        include_leap_day (bool, optional): If False, remove data from leap day if the
+            resource_year is a leap year. Otherwise, leave leap day data in. Defaults to False.
         resource_data (dict | object, optional): Dictionary of user-input resource data.
             Defaults to an empty dictionary.
         resource_dir (str | Path, optional): Folder to save resource files to or
@@ -42,6 +44,7 @@ class OpenMeteoHistoricalSolarAPIConfig(ResourceBaseAPIConfig):
     """
 
     resource_year: int = field(converter=int, validator=range_val(1940, datetime.now().year - 1))
+    include_leap_day: bool = field(default=False)
     dataset_desc: str = "openmeteo_archive_solar"
     resource_type: str = "solar"
     valid_intervals: list[int] = field(factory=lambda: [60])
@@ -286,7 +289,41 @@ def load_data(self, fpath):
         data["Minute"] = time.minute
 
         data = data[data["Year"] == self.config.resource_year]
-        # TODO: throw error if data isn't proper length
+
+        data = data.reset_index(drop=True)
+
+        # Check if data includes leap day
+        data_has_leap_day = int(data[data["Month"] == 2]["Day"].max()) == 29
+
+        # Remove leap day if needed
+        if not self.config.include_leap_day and data_has_leap_day:
+            # Get index of dataframe that includes leap day
+            leap_day_index = (
+                data.reset_index(drop=False)
+                .set_index(keys=["Month", "Day"], drop=True)
+                .loc[(2, 29)]["index"]
+                .to_list()
+            )
+            # Drop the leap day data from the dataframe
+            data = data.drop(index=leap_day_index)
+
+        # Check if data is the same length as the number of timesteps
+        if len(data) != self.n_timesteps:
+            leap_day_msg = ""
+            if data_has_leap_day and len(data) > self.n_timesteps:
+                # Add extra detail to error message if error may be due to leap day
+                leap_day_msg = (
+                    "This may be because the resource data includes a leap day. ",
+                    "To remove data from a leap day from resource data, please set "
+                    "`include_leap_day` to False.",
+                )
+
+            msg = (
+                f"{self.__class__.__name__}: Resource data is not the same length as n_timesteps. "
+                f"Resource data has length {len(data)}, n_timesteps is {self.n_timesteps}. "
+                f"{leap_day_msg}"
+            )
+            raise ValueError(msg)
 
         data, data_units = self.format_timeseries_data(data)
         # make units for data in openmdao-compatible units

h2integrate/resource/solar/test/test_pvwatts_integration.py

@@ -72,6 +72,7 @@ def pysam_performance_model():
         ("MeteosatPrimeMeridianSolarAPI", "solar", 41.9077, 12.4368, 2008, "nsrdb_msg_v4", 0, 2e-2, 410211.9419),  # noqa: E501
         pytest.param("MeteosatPrimeMeridianTMYSolarAPI", "solar", -27.3649, 152.67935, "tmy-2022", "himawari_tmy_v3", 0, 1e-3, 510709.633402, marks=pytest.mark.xfail(reason="Longitude mismatch")),  # noqa: E501
         ("OpenMeteoHistoricalSolarResource", "solar", 44.04218, -95.19757, 2023, "openmeteo_archive_solar", 0, 1e-3, 443558.17053592583),  # noqa: E501
+        ("OpenMeteoHistoricalSolarResource", "solar", -28.454864, 114.551749, 2024, "openmeteo_archive_solar", 8, 2e-2, 192656.49240723),  # noqa: E501
     ],
     ids=[
         "Himawari7SolarAPI",
@@ -80,6 +81,7 @@ def pysam_performance_model():
         "MeteosatPrimeMeridianSolarAPI",
         "MeteosatPrimeMeridianTMYSolarAPI",
         "OpenMeteoHistoricalSolarResource",
+        "OpenMeteoHistoricalSolarResource-LeapYear",
     ]
 )
 # fmt: on

h2integrate/resource/solar/test/test_resource_models.py

@@ -197,3 +197,61 @@ def test_solar_resource_h2i_download(
         assert solar_data["start_time"] == f"{resource_year}/01/01 00:00:00 (+0000)"
     with subtests.test("Time step"):
         assert solar_data["dt"] == plant_simulation["simulation"]["dt"]
+
+
+
+# fmt: off
+@pytest.mark.unit
+@pytest.mark.parametrize(
+    "model,which,lat,lon,resource_year,model_name,timezone",
+    [("OpenMeteoHistoricalSolarResource", "solar",  -28.454864, 114.551749, 2024, "openmeteo_archive_solar", 8)],  # noqa: E501
+    ids=["OpenMeteoHistoricalSolarResource-LeapYear"]
+)
+# fmt: on
+def test_solar_resource_h2i_download_leap_year(
+    plant_simulation,
+    site_config,
+    subtests,
+    model,
+    which,
+    lat,
+    lon,
+    resource_year,
+    model_name,
+):
+    plant_config = {
+        "site": site_config,
+        "plant": plant_simulation,
+    }
+
+    prob = om.Problem()
+    comp = supported_models[model](
+        plant_config=plant_config,
+        resource_config=plant_config["site"]["resources"]["solar_resource"]["resource_parameters"],
+        driver_config={},
+    )
+    prob.model.add_subsystem("resource", comp)
+    prob.setup()
+    prob.run_model()
+    solar_data = prob.get_val("resource.solar_resource_data")
+    name_expected = f"{lat}_{lon}_{resource_year}_{model_name}_60min_local_tz.csv"
+    with subtests.test("filepath for data was found where expected"):
+        assert Path(solar_data["filepath"]).exists()
+        assert Path(solar_data["filepath"]).name == name_expected
+
+    with subtests.test("Data timezone"):
+        assert pytest.approx(solar_data["data_tz"], rel=1e-6) == 8
+    with subtests.test("Site Elevation"):
+        assert pytest.approx(solar_data["elevation"], rel=1e-6) == 71
+
+    data_keys = [k for k, v in solar_data.items() if not isinstance(v, float | int | str)]
+    for k in data_keys:
+        with subtests.test(f"{k} resource data is 8760"):
+            assert len(solar_data[k]) == 8760
+
+    with subtests.test("There are 16 timeseries data keys"):
+        assert len(data_keys) == 16
+    with subtests.test("Start time"):
+        assert solar_data["start_time"] == f"{resource_year}/01/01 00:00:00 (+0800)"
+    with subtests.test("Time step"):
+        assert solar_data["dt"] == plant_simulation["simulation"]["dt"]

h2integrate/resource/wind/openmeteo_wind.py

@@ -22,6 +22,8 @@ class OpenMeteoHistoricalWindAPIConfig(ResourceBaseAPIConfig):
     Args:
         resource_year (int): Year to use for resource data.
             Must been between 1940 the year before the current calendar year. (inclusive).
+        include_leap_day (bool, optional): If False, remove data from leap day if the
+            resource_year is a leap year. Otherwise, leave leap day data in. Defaults to False.
         verify_download (bool, optional): Whether to verify the API download from the url.
             If an `openmeteo_requests.Client.OpenMeteoRequestsError` error is thrown,
             try setting to True. Defaults to False.
@@ -37,6 +39,7 @@ class OpenMeteoHistoricalWindAPIConfig(ResourceBaseAPIConfig):
     """
 
     resource_year: int = field(converter=int, validator=range_val(1940, datetime.now().year - 1))
+    include_leap_day: bool = field(default=False)
     dataset_desc: str = "openmeteo_archive"
     resource_type: str = "wind"
     valid_intervals: list[int] = field(factory=lambda: [60])
@@ -279,7 +282,41 @@ def load_data(self, fpath):
         data["Minute"] = time.minute
 
         data = data[data["Year"] == self.config.resource_year]
-        # TODO: throw error if data isn't proper length
+
+        data = data.reset_index(drop=True)
+
+        # Check if data includes leap day
+        data_has_leap_day = int(data[data["Month"] == 2]["Day"].max()) == 29
+
+        # Remove leap day if needed
+        if not self.config.include_leap_day and data_has_leap_day:
+            # Get index of dataframe that includes leap day
+            leap_day_index = (
+                data.reset_index(drop=False)
+                .set_index(keys=["Month", "Day"], drop=True)
+                .loc[(2, 29)]["index"]
+                .to_list()
+            )
+            # Drop the leap day data from the dataframe
+            data = data.drop(index=leap_day_index)
+
+        # Check if data is the same length as the number of timesteps
+        if len(data) != self.n_timesteps:
+            leap_day_msg = ""
+            if data_has_leap_day and len(data) > self.n_timesteps:
+                # Add extra detail to error message if error may be due to leap day
+                leap_day_msg = (
+                    "This may be because the resource data includes a leap day. ",
+                    "To remove data from a leap day from resource data, please set "
+                    "`include_leap_day` to False.",
+                )
+
+            msg = (
+                f"{self.__class__.__name__}: Resource data is not the same length as n_timesteps. "
+                f"Resource data has length {len(data)}, n_timesteps is {self.n_timesteps}. "
+                f"{leap_day_msg}"
+            )
+            raise ValueError(msg)
 
         data, data_units = self.format_timeseries_data(data)
         # make units for data in openmdao-compatible units
@@ -325,7 +362,7 @@ def format_timeseries_data(self, data):
 
             if "is_day" in c:
                 data_rename_mapper.update({c: "is_day"})
-                data_units.update({"is_day": "percent"})
+                data_units.update({"is_day": "unitless"})
 
             if "surface" in c:
                 new_c += "_0m"

h2integrate/resource/wind/test/test_openmeteo_wind_api.py

@@ -127,3 +127,63 @@ def test_wind_resource_h2i_download(
         assert wind_data["start_time"] == f"{resource_year}/01/01 00:00:00 (-0600)"
     with subtests.test("Time step"):
         assert wind_data["dt"] == plant_simulation["simulation"]["dt"]
+
+
+
+# fmt: off
+@pytest.mark.unit
+@pytest.mark.parametrize(
+    "model,which,lat,lon,resource_year,model_name,timezone,elevation",
+    [("OpenMeteoHistoricalWindResource", "wind", -28.454864, 114.551749, 2024, "openmeteo_archive", 8, 71.0)],  # noqa: E501
+    ids=["Non-UTC Leap Year"],
+)
+# fmt: on
+def test_wind_resource_h2i_download_leap_year(
+    plant_simulation,
+    site_config,
+    subtests,
+    model,
+    lat,
+    lon,
+    resource_year,
+    model_name,
+    timezone,
+    elevation,
+):
+    plant_config = {
+        "site": site_config,
+        "plant": plant_simulation,
+    }
+
+    prob = om.Problem()
+    comp = supported_models[model](
+        plant_config=plant_config,
+        resource_config=plant_config["site"]["resources"]["wind_resource"]["resource_parameters"],
+        driver_config={},
+    )
+    prob.model.add_subsystem("resource", comp)
+    prob.setup()
+    prob.run_model()
+    wind_data = prob.get_val("resource.wind_resource_data")
+
+    expected_fn = f"{lat}_{lon}_{resource_year}_{model_name}_60min_local_tz.csv"
+    with subtests.test("filepath for data was found where expected"):
+        assert Path(wind_data["filepath"]).exists()
+        assert Path(wind_data["filepath"]).name == expected_fn
+
+    with subtests.test("Data timezone"):
+        assert pytest.approx(wind_data["data_tz"], rel=1e-6) == timezone
+    with subtests.test("Site Elevation"):
+        assert pytest.approx(wind_data["elevation"], rel=1e-6) == elevation
+
+    data_keys = [k for k, v in wind_data.items() if not isinstance(v, float | int | str)]
+    for k in data_keys:
+        with subtests.test(f"{k} resource data is 8760"):
+            assert len(wind_data[k]) == 8760
+
+    with subtests.test("theres 14 timeseries data keys"):
+        assert len(data_keys) == 14
+    with subtests.test("Start time"):
+        assert wind_data["start_time"] == f"{resource_year}/01/01 00:00:00 (+0800)"
+    with subtests.test("Time step"):
+        assert wind_data["dt"] == plant_simulation["simulation"]["dt"]