Merge branch 'main' into fix_some_examples_not_run_by_sphinx

Author: echedey-ls

docs/sphinx/source/whatsnew/v0.15.3.rst

@@ -43,11 +43,14 @@ Requirements
 Maintenance
 ~~~~~~~~~~~
 * Fix some gallery examples not being run. Only gallery examples whose filename begins with ``noplot_`` will be excluded. (:issue:`2790`, :pull:`2792`)
+* Removed unused internal function ``clearsky._is_leap_year`` (:issue:`2768`, :pull:`2813`)
 * Fix documentation builds on runner images lacking link-type timezones. (:issue:`2795`, :pull:`2809`)
 
 
 Contributors
 ~~~~~~~~~~~~
 * Eesh Saxena (:ghuser:`eeshsaxena`)
 * Karl Hill (:ghuser:`karlhillx`)
+* Mathias Aschwanden (:ghuser:`maschwanden`)
 * Yonry Zhu (:ghuser:`yonryzhu`)
+* Darshan Gowda (:ghuser:`dgowdaan-cmyk`)

docs/sphinx/source/whatsnew/v0.16.0.rst

@@ -0,0 +1,50 @@
+.. _whatsnew_0_16_0:
+
+
+v0.16.0
+Breaking Changes
+~~~~~~~~~~~~~~~~
+* Change output type of :py:func:`pvlib.irradiance.perez` and
+  :py:func:`pvlib.irradiance.perez_driesse` from ``OrderedDict`` to ``dict`` when
+  ``return_components=True`` to be consistent with other models.
+  (:pull:`2789`)
+* Remove empty ``poa_horizon`` key from the ``diffuse_components`` output of
+  :py:func:`pvlib.irradiance.haydavies`.
+  (:pull:`2788`)
+
+
+Deprecations
+~~~~~~~~~~~~
+
+
+Bug fixes
+~~~~~~~~~
+
+
+Enhancements
+~~~~~~~~~~~~
+
+
+Documentation
+~~~~~~~~~~~~~
+
+
+Testing
+~~~~~~~
+
+
+Benchmarking
+~~~~~~~~~~~~
+
+
+Requirements
+~~~~~~~~~~~~
+
+
+Maintenance
+~~~~~~~~~~~
+
+
+Contributors
+~~~~~~~~~~~~
+* Carolina Crespo (:ghuser:`cbcrespo`)

pvlib/clearsky.py

@@ -220,22 +220,6 @@ def lookup_linke_turbidity(time, latitude, longitude, filepath=None,
     return linke_turbidity
 
 
-def _is_leap_year(year):
-    """Determine if a year is leap year.
-
-    Parameters
-    ----------
-    year : numeric
-
-    Returns
-    -------
-    isleap : array of bools
-    """
-    isleap = ((np.mod(year, 4) == 0) &
-              ((np.mod(year, 100) != 0) | (np.mod(year, 400) == 0)))
-    return isleap
-
-
 def _interpolate_turbidity(lts, time):
     """
     Interpolated monthly Linke turbidity onto daily values.

pvlib/irradiance.py

@@ -787,7 +787,7 @@ def haydavies(surface_tilt, surface_azimuth, dhi, dni, dni_extra,
 
     Returns
     --------
-    numeric, OrderedDict, or DataFrame
+    numeric, dict, or DataFrame
         Return type controlled by ``return_components`` argument.
         If `False`, ``sky_diffuse`` is returned.
         If `True`, ``diffuse_components`` is returned.
@@ -796,13 +796,11 @@ def haydavies(surface_tilt, surface_azimuth, dhi, dni, dni_extra,
         The sky diffuse component of the solar radiation on a tilted
         surface. [Wm⁻²]
 
-    diffuse_components : OrderedDict (array input) or DataFrame (Series input)
+    diffuse_components : dict (array input) or DataFrame (Series input)
         Keys/columns are:
             * poa_sky_diffuse: Total sky diffuse
             * poa_isotropic
             * poa_circumsolar
-            * poa_horizon (always zero, not accounted for by the
-              Hay-Davies model)
 
     Notes
     ------
@@ -860,14 +858,11 @@ def haydavies(surface_tilt, surface_azimuth, dhi, dni, dni_extra,
     sky_diffuse = poa_isotropic + poa_circumsolar
 
     if return_components:
-        diffuse_components = OrderedDict()
-        diffuse_components['poa_sky_diffuse'] = sky_diffuse
-
-        # Calculate the individual components
-        diffuse_components['poa_isotropic'] = poa_isotropic
-        diffuse_components['poa_circumsolar'] = poa_circumsolar
-        diffuse_components['poa_horizon'] = np.where(
-            np.isnan(diffuse_components['poa_isotropic']), np.nan, 0.)
+        diffuse_components = {
+            'poa_sky_diffuse': sky_diffuse,
+            'poa_isotropic': poa_isotropic,
+            'poa_circumsolar': poa_circumsolar
+        }
 
         if isinstance(sky_diffuse, pd.Series):
             diffuse_components = pd.DataFrame(diffuse_components)
@@ -1109,7 +1104,7 @@ def perez(surface_tilt, surface_azimuth, dhi, dni, dni_extra,
 
     Returns
     --------
-    numeric, OrderedDict, or DataFrame
+    numeric, dict, or DataFrame
         Return type controlled by `return_components` argument.
         If ``return_components=False``, `sky_diffuse` is returned.
         If ``return_components=True``, `diffuse_components` is returned.
@@ -1118,7 +1113,7 @@ def perez(surface_tilt, surface_azimuth, dhi, dni, dni_extra,
         The sky diffuse component of the solar radiation on a tilted
         surface.
 
-    diffuse_components : OrderedDict (array input) or DataFrame (Series input)
+    diffuse_components : dict (array input) or DataFrame (Series input)
         Keys/columns are:
             * poa_sky_diffuse: Total sky diffuse
             * poa_isotropic
@@ -1205,13 +1200,12 @@ def perez(surface_tilt, surface_azimuth, dhi, dni, dni_extra,
         sky_diffuse = np.where(np.isnan(airmass), 0, sky_diffuse)
 
     if return_components:
-        diffuse_components = OrderedDict()
-        diffuse_components['poa_sky_diffuse'] = sky_diffuse
-
-        # Calculate the different components
-        diffuse_components['poa_isotropic'] = dhi * term1
-        diffuse_components['poa_circumsolar'] = dhi * term2
-        diffuse_components['poa_horizon'] = dhi * term3
+        diffuse_components = {
+            'poa_sky_diffuse': sky_diffuse,
+            'poa_isotropic': dhi * term1,
+            'poa_circumsolar': dhi * term2,
+            'poa_horizon': dhi * term3
+        }
 
         # Set values of components to 0 when sky_diffuse is 0
         mask = sky_diffuse == 0
@@ -1351,7 +1345,7 @@ def perez_driesse(surface_tilt, surface_azimuth, dhi, dni, dni_extra,
 
     Returns
     --------
-    numeric, OrderedDict, or DataFrame
+    numeric, dict, or DataFrame
         Return type controlled by `return_components` argument.
         If ``return_components=False``, `sky_diffuse` is returned.
         If ``return_components=True``, `diffuse_components` is returned.
@@ -1360,7 +1354,7 @@ def perez_driesse(surface_tilt, surface_azimuth, dhi, dni, dni_extra,
         The sky diffuse component of the solar radiation on a tilted
         surface.
 
-    diffuse_components : OrderedDict (array input) or DataFrame (Series input)
+    diffuse_components : dict (array input) or DataFrame (Series input)
         Keys/columns are:
             * poa_sky_diffuse: Total sky diffuse
             * poa_isotropic
@@ -1425,13 +1419,12 @@ def perez_driesse(surface_tilt, surface_azimuth, dhi, dni, dni_extra,
     sky_diffuse = np.maximum(dhi * (term1 + term2 + term3), 0)
 
     if return_components:
-        diffuse_components = OrderedDict()
-        diffuse_components['poa_sky_diffuse'] = sky_diffuse
-
-        # Calculate the different components
-        diffuse_components['poa_isotropic'] = dhi * term1
-        diffuse_components['poa_circumsolar'] = dhi * term2
-        diffuse_components['poa_horizon'] = dhi * term3
+        diffuse_components = {
+            'poa_sky_diffuse': sky_diffuse,
+            'poa_isotropic': dhi * term1,
+            'poa_circumsolar': dhi * term2,
+            'poa_horizon': dhi * term3
+        }
 
         if isinstance(sky_diffuse, pd.Series):
             diffuse_components = pd.DataFrame(diffuse_components)

tests/iotools/test_meteonorm.py

@@ -61,30 +61,6 @@ def expected_meteonorm_index():
     return expected_meteonorm_index
 
 
-@pytest.fixture
-def expected_meteonorm_data():
-    # The first 12 rows of data
-    columns = ['ghi', 'global_horizontal_irradiance_with_shading']
-    expected = [
-        [0.0, 0.0],
-        [0.0, 0.0],
-        [0.0, 0.0],
-        [0.0, 0.0],
-        [0.0, 0.0],
-        [0.0, 0.0],
-        [0.0, 0.0],
-        [3.75, 3.74],
-        [57.25, 57.20],
-        [149.0, 148.96],
-        [242.25, 242.24],
-        [228.0, 227.98],
-    ]
-    index = pd.date_range('2025-01-01 00:30', periods=12, freq='1h', tz='UTC')
-    index.freq = None
-    expected = pd.DataFrame(expected, index=index, columns=columns)
-    return expected
-
-
 @pytest.fixture
 def expected_columns_all():
     columns = [
@@ -114,8 +90,7 @@ def expected_columns_all():
 @pytest.mark.remote_data
 @pytest.mark.flaky(reruns=RERUNS, reruns_delay=RERUNS_DELAY)
 def test_get_meteonorm_training(
-        demo_api_key, demo_url, expected_meta, expected_meteonorm_index,
-        expected_meteonorm_data):
+        demo_api_key, demo_url, expected_meta, expected_meteonorm_index):
     data, meta = pvlib.iotools.get_meteonorm_observation_training(
         latitude=50, longitude=10,
         start='2025-01-01', end='2026-01-01',
@@ -128,10 +103,12 @@ def test_get_meteonorm_training(
     for key in ['version', 'commit']:
         assert key in meta  # value changes, so only check presence
     pd.testing.assert_index_equal(data.index, expected_meteonorm_index)
-    # meteonorm API only guarantees similar, not identical, results between
-    # calls.  so we allow a small amount of variation with atol.
-    pd.testing.assert_frame_equal(data.iloc[:12], expected_meteonorm_data,
-                                  check_exact=False, atol=1)
+    # don't pin values: meteonorm may update the dataset without it being a
+    # breaking change.  check parsing instead.
+    assert list(data.columns) == \
+        ['ghi', 'global_horizontal_irradiance_with_shading']
+    assert data.dtypes.map(pd.api.types.is_numeric_dtype).all()
+    assert (data.isna().mean() <= 0.2).all()  # meteonorm guarantees <=20% NaN
 
 
 @pytest.mark.remote_data
@@ -156,8 +133,11 @@ def test_get_meteonorm_realtime(demo_api_key, demo_url, expected_columns_all):
     assert meta['surface_tilt'] == 20
     assert meta['surface_azimuth'] == 10
 
-    assert list(data.columns) == expected_columns_all
-    assert data.shape == (241, 19)
+    # meteonorm may add parameters to 'all' at any time, so only check that
+    # the columns we know about are present, not that the set matches exactly.
+    assert set(expected_columns_all).issubset(data.columns)
+    assert data.shape[0] == 241  # row count is determined by the time range
+    assert data.shape[1] >= len(expected_columns_all)
     # can't test the specific index as it varies due to the
     # use of pd.Timestamp.now
     assert type(data.index) is pd.core.indexes.interval.IntervalIndex
@@ -259,38 +239,10 @@ def expected_meteonorm_tmy_meta():
     return meta
 
 
-@pytest.fixture
-def expected_meteonorm_tmy_data():
-    # The first 12 rows of data
-    columns = ['diffuse_horizontal_irradiance']
-    expected = [
-        [0.],
-        [0.],
-        [0.],
-        [0.],
-        [0.],
-        [0.],
-        [0.],
-        [0.],
-        [9.07],
-        [8.44],
-        [86.64],
-        [110.44],
-    ]
-    index = pd.date_range(
-        '2030-01-01', periods=12, freq='1h', tz=3600)
-    index.freq = None
-    interval_index = pd.IntervalIndex.from_arrays(
-        index, index + pd.Timedelta(hours=1), closed='left')
-    expected = pd.DataFrame(expected, index=interval_index, columns=columns)
-    return expected
-
-
 @pytest.mark.remote_data
 @pytest.mark.flaky(reruns=RERUNS, reruns_delay=RERUNS_DELAY)
 def test_get_meteonorm_tmy(
-        demo_api_key, demo_url, expected_meteonorm_tmy_meta,
-        expected_meteonorm_tmy_data):
+        demo_api_key, demo_url, expected_meteonorm_tmy_meta):
     data, meta = pvlib.iotools.get_meteonorm_tmy(
         latitude=50, longitude=10,
         api_key=demo_api_key,
@@ -312,10 +264,11 @@ def test_get_meteonorm_tmy(
     assert meta.items() >= expected_meteonorm_tmy_meta.items()
     for key in ['version', 'commit']:
         assert key in meta  # value changes, so only check presence
-    # meteonorm API only guarantees similar, not identical, results between
-    # calls.  so we allow a small amount of variation with atol.
-    pd.testing.assert_frame_equal(data.iloc[:12], expected_meteonorm_tmy_data,
-                                  check_exact=False, atol=1)
+    # don't pin values: meteonorm may update the dataset without it being a
+    # breaking change.  check parsing instead.
+    assert list(data.columns) == ['diffuse_horizontal_irradiance']
+    assert data.dtypes.map(pd.api.types.is_numeric_dtype).all()
+    assert (data.isna().mean() <= 0.2).all()  # meteonorm guarantees <=20% NaN
 
 
 @fail_on_pvlib_version('0.16.0')

tests/test_irradiance.py

@@ -209,13 +209,11 @@ def test_haydavies(irrad_data, ephem_data, dni_et):
 
 
 def test_haydavies_components(irrad_data, ephem_data, dni_et):
-    keys = ['poa_sky_diffuse', 'poa_isotropic', 'poa_circumsolar',
-            'poa_horizon']
+    keys = ['poa_sky_diffuse', 'poa_isotropic', 'poa_circumsolar']
     expected = pd.DataFrame(np.array(
         [[0, 27.1775, 102.9949, 33.1909],
          [0, 27.1775, 30.1818, 27.9837],
-         [0, 0, 72.8130, 5.2071],
-         [0, 0, 0, 0]]).T,
+         [0, 0, 72.8130, 5.2071]]).T,
         columns=keys,
         index=irrad_data.index
     )