mypy compliance

CHANGES.rst

@@ -25,7 +25,7 @@ Breaking changes
 * The previously deprecated function ``xclim.ensembles.change_significance`` has been removed. (:pull:`1737`).
 * Indicators ``snw_season_length`` and ``snd_season_length`` have been modified, see above.
 * The `hargeaves85`/`hg85` method for the ``potential_evapotranspiration`` indicator and indice has been modified for precision and consistency with recent academic literature. (:issue:`1710`, :pull:`1723`).
-
+* The `__getitem__` method of ``xclim.core.indicator.Parameter`` instances has been removed. Accessing members of ``Parameters`` now uniquely uses dot notation. (:pull:`1721`).
 
 Bug fixes
 ^^^^^^^^^
@@ -36,6 +36,8 @@ Bug fixes
 * Fixed "agreement fraction" in ``robustness_fractions`` to distinguish between negative change and no change. Added "negative" and "changed negative" fractions (:issue:`1690`, :pull:`1711`).
 * ``make_criteria`` now skips columns with NaNs across all realizations. (:pull:`1713`).
 * Fixed bug QuantileDeltaMapping adjustment not working for seasonal grouping (:issue:`1704`, :pull:`1716`).
+* The codebase has been adjusted to address several (~300) `mypy`-related typing errors. (:issue:`1719`, :pull:`1721`).
+* PEP8 rule `N802` is now enabled in the `ruff` formatter. (:pull:`1721`).
 
 Internal changes
 ^^^^^^^^^^^^^^^^

environment.yml

@@ -53,6 +53,7 @@ dependencies:
   - nc-time-axis
   - netCDF4 >=1.4
   - notebook
+  - pandas-stubs
   - platformdirs
   - pooch
   - pre-commit

pyproject.toml

@@ -71,6 +71,7 @@ dev = [
   "nbqa",
   "nbval",
   "netCDF4 >=1.4",
+  "pandas-stubs>=2.2",
   "platformdirs >=3.2",
   "pre-commit >=2.9",
   "pybtex",
@@ -207,24 +208,20 @@ python_version = 3.9
 show_error_codes = true
 warn_return_any = true
 warn_unused_configs = true
+plugins = ["numpy.typing.mypy_plugin"]
 
 [[tool.mypy.overrides]]
 module = [
   "boltons.*",
-  "bottleneck.*",
   "cftime.*",
-  "clisops.core.subset.*",
-  "dask.*",
-  "lmoments3.*",
-  "matplotlib.*",
+  "jsonpickle.*",
   "numba.*",
-  "numpy.*",
-  "pandas.*",
-  "pint.*",
+  "pytest_socket.*",
   "SBCK.*",
   "scipy.*",
-  "sklearn.cluster.*",
-  "xarray.*",
+  "sklearn.*",
+  "statsmodels.*",
+  "yamale.*",
   "yaml.*"
 ]
 ignore_missing_imports = true
@@ -273,6 +270,7 @@ select = [
   "D",
   "E",
   "F",
+  "N802",
   "W"
 ]
 

tests/test_analog.py

@@ -196,7 +196,7 @@ def test_compare_with_matlab(self):
 
 
 class TestKS:
-    def test_1D_ks_2samp(self, random):
+    def test_1D_ks_2samp(self, random):  # noqa: N802
         # Compare with scipy.stats.ks_2samp
         x = random.standard_normal(50) + 1
         y = random.standard_normal(50)
@@ -210,7 +210,7 @@ def test_compare_with_matlab(self):
         assert_almost_equal(dm, 0.96667, 4)
 
 
-def analytical_KLDiv(p, q):
+def analytical_KLDiv(p, q):  # noqa: N802
     """Return the Kullback-Leibler divergence between two distributions.
 
     Parameters

tests/test_cffwis.py

@@ -389,7 +389,7 @@ def test_fire_weather_ufunc_errors(
             ),
         ],
     )
-    def test_fire_season_R(self, open_dataset, key, kwargs):
+    def test_fire_season_R(self, open_dataset, key, kwargs):  # noqa: N802
         expected = self._get_cffdrs_fire_season(key)
         in_ds = open_dataset("FWI/cffdrs_test_wDC.nc")
         nid = int(key[2])

tests/test_indicators.py

@@ -225,7 +225,7 @@ def test_opt_vars(tasmin_series, tasmax_series):
     tx = tasmax_series(np.zeros(365))
 
     multiOptVar(tasmin=tn, tasmax=tx)
-    assert multiOptVar.parameters["tasmin"]["kind"] == InputKind.OPTIONAL_VARIABLE
+    assert multiOptVar.parameters["tasmin"].kind == InputKind.OPTIONAL_VARIABLE
 
 
 def test_registering():
@@ -265,8 +265,10 @@ def test_module():
     """Translations are keyed according to the module where the indicators are defined."""
     assert atmos.tg_mean.__module__.split(".")[2] == "atmos"
     # Virtual module also are stored under xclim.indicators
-    assert xclim.indicators.cf.fg.__module__ == "xclim.indicators.cf"
-    assert xclim.indicators.icclim.GD4.__module__ == "xclim.indicators.icclim"
+    assert xclim.indicators.cf.fg.__module__ == "xclim.indicators.cf"  # noqa: F821
+    assert (
+        xclim.indicators.icclim.GD4.__module__ == "xclim.indicators.icclim"
+    )  # noqa: F821
 
 
 def test_temp_unit_conversion(tas_series):
@@ -377,7 +379,7 @@ def test_multiindicator(tas_series):
         compute=uniindtemp_compute,
     )
     with pytest.raises(ValueError, match="Indicator minmaxtemp4 was wrongly defined"):
-        tmin, tmax = ind(tas, freq="YS")
+        _tmin, _tmax = ind(tas, freq="YS")
 
 
 def test_missing(tas_series):
@@ -480,7 +482,7 @@ def test_all_parameters_understood(official_indicators):
     for identifier, ind in official_indicators.items():
         indinst = ind.get_instance()
         for name, param in indinst.parameters.items():
-            if param["kind"] == InputKind.OTHER_PARAMETER:
+            if param.kind == InputKind.OTHER_PARAMETER:
                 problems.add((identifier, name))
     # this one we are ok with.
     if problems - {
@@ -587,15 +589,15 @@ def test_parsed_doc():
     assert "tas" in xclim.atmos.liquid_precip_accumulation.parameters
 
     params = xclim.atmos.drought_code.parameters
-    assert params["tas"]["description"] == "Noon temperature."
-    assert params["tas"]["units"] == "[temperature]"
-    assert params["tas"]["kind"] is InputKind.VARIABLE
-    assert params["tas"]["default"] == "tas"
-    assert params["snd"]["default"] is None
-    assert params["snd"]["kind"] is InputKind.OPTIONAL_VARIABLE
-    assert params["snd"]["units"] == "[length]"
-    assert params["season_method"]["kind"] is InputKind.STRING
-    assert params["season_method"]["choices"] == {"GFWED", None, "WF93", "LA08"}
+    assert params["tas"].description == "Noon temperature."
+    assert params["tas"].units == "[temperature]"
+    assert params["tas"].kind is InputKind.VARIABLE
+    assert params["tas"].default == "tas"
+    assert params["snd"].default is None
+    assert params["snd"].kind is InputKind.OPTIONAL_VARIABLE
+    assert params["snd"].units == "[length]"
+    assert params["season_method"].kind is InputKind.STRING
+    assert params["season_method"].choices == {"GFWED", None, "WF93", "LA08"}
 
 
 def test_default_formatter():
@@ -604,7 +606,7 @@ def test_default_formatter():
     assert default_formatter.format("{month}", month="m3") == "march"
 
 
-def test_AttrFormatter():
+def test_AttrFormatter():  # noqa: N802
     fmt = AttrFormatter(
         mapping={"evil": ["méchant", "méchante"], "nice": ["beau", "belle"]},
         modifiers=["m", "f"],
@@ -655,14 +657,14 @@ def test_input_dataset(open_dataset):
     ds = open_dataset("ERA5/daily_surface_cancities_1990-1993.nc")
 
     # Use defaults
-    out = xclim.atmos.daily_temperature_range(freq="YS", ds=ds)
+    _ = xclim.atmos.daily_temperature_range(freq="YS", ds=ds)
 
     # Use non-defaults (inverted on purpose)
     with xclim.set_options(cf_compliance="log"):
-        out = xclim.atmos.daily_temperature_range("tasmax", "tasmin", freq="YS", ds=ds)
+        _ = xclim.atmos.daily_temperature_range("tasmax", "tasmin", freq="YS", ds=ds)
 
     # Use a mix
-    out = xclim.atmos.daily_temperature_range(tasmax=ds.tasmax, freq="YS", ds=ds)
+    _ = xclim.atmos.daily_temperature_range(tasmax=ds.tasmax, freq="YS", ds=ds)
 
     # Inexistent variable:
     dsx = ds.drop_vars("tasmin")
@@ -817,7 +819,7 @@ def test_indicator_call_errors(tas_series):
         uniIndTemp(tas, oups=3)
 
 
-def test_resamplingIndicator_new_error():
+def test_resamplingIndicator_new_error():  # noqa: N802
     with pytest.raises(ValueError, match="ResamplingIndicator require a 'freq'"):
         Daily(
             realm="atmos",

tests/test_indices.py

@@ -3717,42 +3717,42 @@ def test_wet_spell_frequency_op(pr_series):
 
 
 class TestSfcWindMax:
-    def test_sfcWind_max(self, sfcWind_series):
+    def test_sfcWind_max(self, sfcWind_series):  # noqa: N802
         sfcWind = sfcWind_series(np.array([14.11, 15.27, 10.70]))
         out = xci.sfcWind_max(sfcWind)
         np.testing.assert_allclose(out, [15.27])
 
 
 class TestSfcWindMean:
-    def test_sfcWind_mean(self, sfcWind_series):
+    def test_sfcWind_mean(self, sfcWind_series):  # noqa: N802
         sfcWind = sfcWind_series(np.array([14.11, 15.27, 10.70]))
         out = xci.sfcWind_mean(sfcWind)
         np.testing.assert_allclose(out, [13.36])
 
 
 class TestSfcWindMin:
-    def test_sfcWind_min(self, sfcWind_series):
+    def test_sfcWind_min(self, sfcWind_series):  # noqa: N802
         sfcWind = sfcWind_series(np.array([14.11, 15.27, 10.70]))
         out = xci.sfcWind_min(sfcWind)
         np.testing.assert_allclose(out, [10.70])
 
 
 class TestSfcWindmaxMax:
-    def test_sfcWindmax_max(self, sfcWindmax_series):
+    def test_sfcWindmax_max(self, sfcWindmax_series):  # noqa: N802
         sfcWindmax = sfcWindmax_series(np.array([14.11, 15.27, 10.70]))
         out = xci.sfcWindmax_max(sfcWindmax)
         np.testing.assert_allclose(out, [15.27])
 
 
 class TestSfcWindmaxMean:
-    def test_sfcWindmax_mean(self, sfcWindmax_series):
+    def test_sfcWindmax_mean(self, sfcWindmax_series):  # noqa: N802
         sfcWindmax = sfcWindmax_series(np.array([14.11, 15.27, 10.70]))
         out = xci.sfcWindmax_mean(sfcWindmax)
         np.testing.assert_allclose(out, [13.36])
 
 
 class TestSfcWindmaxMin:
-    def test_sfcWindmax_min(self, sfcWindmax_series):
+    def test_sfcWindmax_min(self, sfcWindmax_series):  # noqa: N802
         sfcWindmax = sfcWindmax_series(np.array([14.11, 15.27, 10.70]))
         out = xci.sfcWindmax_min(sfcWindmax)
         np.testing.assert_allclose(out, [10.70])

tests/test_sdba/test_adjustment.py

@@ -101,7 +101,9 @@ def test_time(self, kind, name, series, random):
         np.testing.assert_array_almost_equal(p, ref)
 
     @pytest.mark.parametrize("kind,name", [(ADDITIVE, "tas"), (MULTIPLICATIVE, "pr")])
-    def test_mon_U(self, mon_series, series, mon_triangular, kind, name, random):
+    def test_mon_U(  # noqa: N802
+        self, mon_series, series, mon_triangular, kind, name, random
+    ):
         n = 10000
         u = random.random(n)
 
@@ -209,7 +211,9 @@ def test_quantiles(self, series, kind, name, random):
 
     @pytest.mark.parametrize("kind,name", [(ADDITIVE, "tas"), (MULTIPLICATIVE, "pr")])
     @pytest.mark.parametrize("add_dims", [True, False])
-    def test_mon_U(self, mon_series, series, kind, name, add_dims, random):
+    def test_mon_U(  # noqa: N802
+        self, mon_series, series, kind, name, add_dims, random
+    ):
         """
         Train on
         hist: U
@@ -324,7 +328,7 @@ def test_quantiles(self, series, kind, name, random):
     @pytest.mark.parametrize("use_dask", [True, False])
     @pytest.mark.parametrize("kind,name", [(ADDITIVE, "tas"), (MULTIPLICATIVE, "pr")])
     @pytest.mark.parametrize("add_dims", [True, False])
-    def test_mon_U(
+    def test_mon_U(  # noqa: N802
         self, mon_series, series, mon_triangular, add_dims, kind, name, use_dask, random
     ):
         """
@@ -480,7 +484,9 @@ def test_quantiles(self, series, kind, name, random):
         np.testing.assert_array_almost_equal(p[middle], ref[middle], 1)
 
     @pytest.mark.parametrize("kind,name", [(ADDITIVE, "tas"), (MULTIPLICATIVE, "pr")])
-    def test_mon_U(self, mon_series, series, mon_triangular, kind, name, random):
+    def test_mon_U(  # noqa: N802
+        self, mon_series, series, mon_triangular, kind, name, random
+    ):
         """
         Train on
         hist: U

tests/test_temperature.py

@@ -69,7 +69,7 @@ class TestDTR:
     nc_tasmax = os.path.join("NRCANdaily", "nrcan_canada_daily_tasmax_1990.nc")
     nc_tasmin = os.path.join("NRCANdaily", "nrcan_canada_daily_tasmin_1990.nc")
 
-    def test_DTR_3d_data_with_nans(self, open_dataset):
+    def test_DTR_3d_data_with_nans(self, open_dataset):  # noqa: N802
         tasmax = open_dataset(self.nc_tasmax).tasmax
         tasmax_C = open_dataset(self.nc_tasmax).tasmax
         tasmax_C -= K2C
@@ -176,7 +176,7 @@ class TestTmean:
         os.path.join("NRCANdaily", "nrcan_canada_daily_tasmin_1990.nc"),
     )
 
-    def test_Tmean_3d_data(self, open_dataset):
+    def test_Tmean_3d_data(self, open_dataset):  # noqa: N802
         ds_tmax = open_dataset(self.nc_files[0])
         ds_tmin = open_dataset(self.nc_files[1])
         tas = atmos.tg(ds_tmin.tasmin, ds_tmax.tasmax)
@@ -205,7 +205,7 @@ def test_Tmean_3d_data(self, open_dataset):
 class TestTx:
     nc_file = os.path.join("NRCANdaily", "nrcan_canada_daily_tasmax_1990.nc")
 
-    def test_TX_3d_data(self, open_dataset):
+    def test_TX_3d_data(self, open_dataset):  # noqa: N802
         tasmax = open_dataset(self.nc_file).tasmax
         tasmax_C = open_dataset(self.nc_file).tasmax
         tasmax_C.values -= K2C
@@ -255,7 +255,7 @@ def test_TX_3d_data(self, open_dataset):
 class TestTn:
     nc_file = os.path.join("NRCANdaily", "nrcan_canada_daily_tasmin_1990.nc")
 
-    def test_TN_3d_data(self, open_dataset):
+    def test_TN_3d_data(self, open_dataset):  # noqa: N802
         tasmin = open_dataset(self.nc_file).tasmin
         tasmin_C = open_dataset(self.nc_file).tasmin
         tasmin_C.values -= K2C

tests/test_utils.py

@@ -2,8 +2,6 @@
 # Test for utils
 from __future__ import annotations
 
-from inspect import signature
-
 import numpy as np
 import xarray as xr
 
@@ -12,7 +10,6 @@
     ensure_chunk_size,
     nan_calc_percentiles,
     walk_map,
-    wrapped_partial,
 )
 from xclim.testing.helpers import test_timeseries as _test_timeseries
 
@@ -24,29 +21,6 @@ def test_walk_map():
     assert o["b"]["c"] == 0
 
 
-def test_wrapped_partial():
-    def func(a, b=1, c=1):
-        """Docstring"""
-        return (a, b, c)
-
-    newf = wrapped_partial(func, b=2)
-    assert list(signature(newf).parameters.keys()) == ["a", "c"]
-    assert newf(1) == (1, 2, 1)
-
-    newf = wrapped_partial(func, suggested=dict(c=2), b=2)
-    assert list(signature(newf).parameters.keys()) == ["a", "c"]
-    assert newf(1) == (1, 2, 2)
-    assert newf.__doc__ == func.__doc__
-
-    def func(a, b=1, c=1, **kws):  # pylint: disable=function-redefined
-        """Docstring"""
-        return a, b, c
-
-    newf = wrapped_partial(func, suggested=dict(c=2), a=2, b=2)
-    assert list(signature(newf).parameters.keys()) == ["c", "kws"]
-    assert newf() == (2, 2, 2)
-
-
 def test_ensure_chunk_size():
     da = xr.DataArray(np.zeros((20, 21, 20)), dims=("x", "y", "z"))