update tests to use new functions

src/xsdba/testing/registry.txt

@@ -1,3 +1,8 @@
+cmip3/tas.sresb1.giss_model_e_r.run1.atm.da.nc sha256:e709552beeeccafcfe280759edf5477ae5241c698409ca051b0899c16e92c95e
+ERA5/daily_surface_cancities_1990-1993.nc sha256:049d54ace3d229a96cc621189daa3e1a393959ab8d988221cfc7b2acd7ab94b2
+NRCANdaily/nrcan_canada_daily_pr_1990.nc sha256:144479ec7a976cfecb6a10762d128a771356093d72caf5f075508ee86d25a1b0
+NRCANdaily/nrcan_canada_daily_tasmax_1990.nc sha256:84880205b798740e37a102c7f40e595d7a4fde6e35fb737a1ef68b8dad447526
+NRCANdaily/nrcan_canada_daily_tasmin_1990.nc sha256:13d61fc54cdcb4c1617ec777ccbf59575d8fdc24754f914042301bc1b024d7f7
 sdba/CanESM2_1950-2100.nc sha256:b41fe603676e70d16c747ec207eb75ec86a39b665de401dcb23b5969ab3e1b32
 sdba/adjusted_external.nc sha256:ff325c88eca96844bc85863744e4e08bcdf3d257388255636427ad5e11960d2e
 sdba/ahccd_1950-2013.nc sha256:7e9a1f61c1d04ca257b09857a82715f1fa3f0550d77f97b7306d4eaaf0c70239

tests/conftest.py

@@ -1,5 +1,4 @@
 # noqa: D104
-# XC: Many things deactivated, not sure what will be necessary
 from __future__ import annotations
 
 from pathlib import Path

tests/test_adjustment.py

@@ -632,24 +632,24 @@ def test_mon_u(
 
     @pytest.mark.parametrize("use_dask", [True, False])
     @pytest.mark.filterwarnings("ignore::RuntimeWarning")
-    def test_add_dims(self, use_dask, open_dataset):
+    def test_add_dims(self, use_dask, gosset):
         with set_options(sdba_encode_cf=use_dask):
             if use_dask:
                 chunks = {"location": -1}
             else:
                 chunks = None
 
-            dsim = open_dataset(
-                "sdba/CanESM2_1950-2100.nc",
+            dsim = xr.open_dataset(
+                gosset.fetch("sdba/CanESM2_1950-2100.nc"),
                 chunks=chunks,
                 drop_variables=["lat", "lon"],
             ).tasmax
             hist = dsim.sel(time=slice("1981", "2010"))
             sim = dsim.sel(time=slice("2041", "2070"))
 
             ref = (
-                open_dataset(
-                    "sdba/ahccd_1950-2013.nc",
+                xr.open_dataset(
+                    gosset.fetch("sdba/ahccd_1950-2013.nc"),
                     chunks=chunks,
                     drop_variables=["lat", "lon"],
                 )
@@ -678,16 +678,16 @@ class TestMBCn:
     @pytest.mark.parametrize("use_dask", [True, False])
     @pytest.mark.parametrize("group, window", [["time", 1], ["time.dayofyear", 31]])
     @pytest.mark.parametrize("period_dim", [None, "period"])
-    def test_simple(self, open_dataset, use_dask, group, window, period_dim):
+    def test_simple(self, use_dask, group, window, period_dim, gosset):
         group, window, period_dim, use_dask = "time", 1, None, False
         with set_options(sdba_encode_cf=use_dask):
             if use_dask:
                 chunks = {"location": -1}
             else:
                 chunks = None
             ref, dsim = (
-                open_dataset(
-                    f"sdba/{file}",
+                xr.open_dataset(
+                    gosset.fetch(f"sdba/{file}"),
                     chunks=chunks,
                     drop_variables=["lat", "lon"],
                 )
@@ -767,8 +767,10 @@ def _group_assert(ds, dim):
 
     @pytest.mark.parametrize("use_dask", [True, False])
     @pytest.mark.parametrize("pcorient", ["full", "simple"])
-    def test_real_data(self, open_dataset, use_dask, pcorient):
-        atmosds = open_dataset("ERA5/daily_surface_cancities_1990-1993.nc")
+    def test_real_data(self, use_dask, pcorient, gosset):
+        atmosds = xr.open_dataset(
+            gosset.fetch("ERA5/daily_surface_cancities_1990-1993.nc")
+        )
 
         ds0 = xr.Dataset(
             {"tasmax": atmosds.tasmax, "tasmin": atmosds.tasmin, "tas": atmosds.tas}
@@ -862,9 +864,9 @@ def gen_testdata(c, s):
         ).sum()
 
     @pytest.mark.slow
-    def test_real_data(self, open_dataset):
-        dsim = open_dataset("sdba/CanESM2_1950-2100.nc")  # .chunk()
-        dref = open_dataset("sdba/ahccd_1950-2013.nc")  # .chunk()
+    def test_real_data(self, gosset):
+        dsim = xr.open_dataset(gosset.fetch("sdba/CanESM2_1950-2100.nc"))  # .chunk()
+        dref = xr.open_dataset(gosset.fetch("sdba/ahccd_1950-2013.nc"))  # .chunk()
         ref = dref.sel(time=slice("1950", "2009")).pr
         hist = dsim.sel(time=slice("1950", "2009")).pr
         # TODO: Do we want to include standard conversions in xsdba tests?

tests/test_base.py

@@ -237,8 +237,8 @@ def func(ds, *, group, lon=None):
             func(xr.Dataset(dict(da0=da0)), group="time")
 
     @pytest.mark.parametrize("use_dask", [True, False])
-    def test_dataarray_cfencode(self, use_dask):
-        ds = open_dataset("sdba/CanESM2_1950-2100.nc")
+    def test_dataarray_cfencode(self, use_dask, gosset):
+        ds = xr.open_dataset(gosset.fetch("sdba/CanESM2_1950-2100.nc"))
         if use_dask:
             ds = ds.chunk()
 

tests/test_loess.py

@@ -40,9 +40,9 @@ def test_loess_nb(d, f, w, n, dx, exp):
 
 @pytest.mark.slow
 @pytest.mark.parametrize("use_dask", [True, False])
-def test_loess_smoothing(use_dask, open_dataset):
-    tas = open_dataset(
-        "cmip3/tas.sresb1.giss_model_e_r.run1.atm.da.nc",
+def test_loess_smoothing(use_dask, gosset):
+    tas = xr.open_dataset(
+        gosset.fetch("cmip3/tas.sresb1.giss_model_e_r.run1.atm.da.nc"),
         chunks={"lat": 1} if use_dask else None,
     ).tas.isel(lon=0, time=slice(0, 740))
     tas = tas.where(tas.time.dt.dayofyear != 360)  # Put NaNs

tests/test_measures.py

@@ -7,16 +7,32 @@
 from xsdba import measures
 
 
-def test_bias(open_dataset):
-    sim = open_dataset("sdba/CanESM2_1950-2100.nc").sel(time="1950-01-01").tasmax
-    ref = open_dataset("sdba/nrcan_1950-2013.nc").sel(time="1950-01-01").tasmax
+def test_bias(gosset):
+    sim = (
+        xr.open_dataset(gosset.fetch("sdba/CanESM2_1950-2100.nc"))
+        .sel(time="1950-01-01")
+        .tasmax
+    )
+    ref = (
+        xr.open_dataset(gosset.fetch("sdba/nrcan_1950-2013.nc"))
+        .sel(time="1950-01-01")
+        .tasmax
+    )
     test = measures.bias(sim, ref).values
     np.testing.assert_array_almost_equal(test, [[6.430237, 39.088974, 5.2402344]])
 
 
-def test_relative_bias(open_dataset):
-    sim = open_dataset("sdba/CanESM2_1950-2100.nc").sel(time="1950-01-01").tasmax
-    ref = open_dataset("sdba/nrcan_1950-2013.nc").sel(time="1950-01-01").tasmax
+def test_relative_bias(gosset):
+    sim = (
+        xr.open_dataset(gosset.fetch("sdba/CanESM2_1950-2100.nc"))
+        .sel(time="1950-01-01")
+        .tasmax
+    )
+    ref = (
+        xr.open_dataset(gosset.fetch("sdba/nrcan_1950-2013.nc"))
+        .sel(time="1950-01-01")
+        .tasmax
+    )
     test = measures.relative_bias(sim, ref).values
     np.testing.assert_array_almost_equal(test, [[0.02366494, 0.16392256, 0.01920133]])
 
@@ -32,36 +48,62 @@ def test_circular_bias():
     np.testing.assert_array_almost_equal(test, [1, 1, 66, -1, -1, -66])
 
 
-def test_ratio(open_dataset):
-    sim = open_dataset("sdba/CanESM2_1950-2100.nc").sel(time="1950-01-01").tasmax
-    ref = open_dataset("sdba/nrcan_1950-2013.nc").sel(time="1950-01-01").tasmax
+def test_ratio(gosset):
+    sim = (
+        xr.open_dataset(gosset.fetch("sdba/CanESM2_1950-2100.nc"))
+        .sel(time="1950-01-01")
+        .tasmax
+    )
+    ref = (
+        xr.open_dataset(gosset.fetch("sdba/nrcan_1950-2013.nc"))
+        .sel(time="1950-01-01")
+        .tasmax
+    )
     test = measures.ratio(sim, ref).values
     np.testing.assert_array_almost_equal(test, [[1.023665, 1.1639225, 1.0192013]])
 
 
-def test_rmse(open_dataset):
+def test_rmse(gosset):
     sim = (
-        open_dataset("sdba/CanESM2_1950-2100.nc").sel(time=slice("1950", "1953")).tasmax
+        xr.open_dataset(gosset.fetch("sdba/CanESM2_1950-2100.nc"))
+        .sel(time=slice("1950", "1953"))
+        .tasmax
+    )
+    ref = (
+        xr.open_dataset(gosset.fetch("sdba/nrcan_1950-2013.nc"))
+        .sel(time=slice("1950", "1953"))
+        .tasmax
     )
-    ref = open_dataset("sdba/nrcan_1950-2013.nc").sel(time=slice("1950", "1953")).tasmax
     test = measures.rmse(sim, ref).values
     np.testing.assert_array_almost_equal(test, [5.4499755, 18.124086, 12.387193], 4)
 
 
-def test_mae(open_dataset):
+def test_mae(gosset):
     sim = (
-        open_dataset("sdba/CanESM2_1950-2100.nc").sel(time=slice("1950", "1953")).tasmax
+        xr.open_dataset(gosset.fetch("sdba/CanESM2_1950-2100.nc"))
+        .sel(time=slice("1950", "1953"))
+        .tasmax
+    )
+    ref = (
+        xr.open_dataset(gosset.fetch("sdba/nrcan_1950-2013.nc"))
+        .sel(time=slice("1950", "1953"))
+        .tasmax
     )
-    ref = open_dataset("sdba/nrcan_1950-2013.nc").sel(time=slice("1950", "1953")).tasmax
     test = measures.mae(sim, ref).values
     np.testing.assert_array_almost_equal(test, [4.159672, 14.2148, 9.768536], 4)
 
 
-def test_annual_cycle_correlation(open_dataset):
+def test_annual_cycle_correlation(gosset):
     sim = (
-        open_dataset("sdba/CanESM2_1950-2100.nc").sel(time=slice("1950", "1953")).tasmax
+        xr.open_dataset(gosset.fetch("sdba/CanESM2_1950-2100.nc"))
+        .sel(time=slice("1950", "1953"))
+        .tasmax
+    )
+    ref = (
+        xr.open_dataset(gosset.fetch("sdba/nrcan_1950-2013.nc"))
+        .sel(time=slice("1950", "1953"))
+        .tasmax
     )
-    ref = open_dataset("sdba/nrcan_1950-2013.nc").sel(time=slice("1950", "1953")).tasmax
     test = (
         measures.annual_cycle_correlation(sim, ref, window=31)
         .sel(location="Vancouver")
@@ -71,22 +113,26 @@ def test_annual_cycle_correlation(open_dataset):
 
 
 @pytest.mark.slow
-def test_scorr(open_dataset):
-    ref = open_dataset("NRCANdaily/nrcan_canada_daily_tasmin_1990.nc").tasmin
-    sim = open_dataset("NRCANdaily/nrcan_canada_daily_tasmax_1990.nc").tasmax
+def test_scorr(gosset):
+    ref = xr.open_dataset(
+        gosset.fetch("NRCANdaily/nrcan_canada_daily_tasmin_1990.nc")
+    ).tasmin
+    sim = xr.open_dataset(
+        gosset.fetch("NRCANdaily/nrcan_canada_daily_tasmax_1990.nc")
+    ).tasmax
     scorr = measures.scorr(sim.isel(lon=slice(0, 50)), ref.isel(lon=slice(0, 50)))
 
     np.testing.assert_allclose(scorr, [97374.2146243])
 
 
-def test_taylordiagram(open_dataset):
+def test_taylordiagram(gosset):
     sim = (
-        open_dataset("sdba/CanESM2_1950-2100.nc")
+        xr.open_dataset(gosset.fetch("sdba/CanESM2_1950-2100.nc"))
         .sel(time=slice("1950", "1953"), location="Amos")
         .tasmax
     )
     ref = (
-        open_dataset("sdba/nrcan_1950-2013.nc")
+        xr.open_dataset(gosset.fetch("sdba/nrcan_1950-2013.nc"))
         .sel(time=slice("1950", "1953"), location="Amos")
         .tasmax
     )

tests/test_nbutils.py

@@ -10,8 +10,11 @@
 class TestQuantiles:
     @pytest.mark.parametrize("uses_dask", [True, False])
     def test_quantile(self, uses_dask, gosset):
-        file = gosset.fetch("sdba/CanESM2_1950-2100.nc")
-        da = (xr.open_dataset(file).sel(time=slice("1950", "1955")).pr).load()
+        da = (
+            xr.open_dataset(gosset.fetch("sdba/CanESM2_1950-2100.nc"))
+            .sel(time=slice("1950", "1955"))
+            .pr
+        ).load()
         if uses_dask:
             da = da.chunk({"location": 1})
         else:

tests/test_processing.py

@@ -279,13 +279,14 @@ def test_normalize(timeseries, random):
     np.testing.assert_allclose(xp, xp2)
 
 
-def test_stack_variables(open_dataset):
-    ds1 = open_dataset("sdba/CanESM2_1950-2100.nc")
-    ds2 = open_dataset("sdba/ahccd_1950-2013.nc")
+def test_stack_variables(gosset):
+    ds1 = xr.open_dataset(gosset.fetch("sdba/CanESM2_1950-2100.nc"))
+    ds2 = xr.open_dataset(gosset.fetch("sdba/ahccd_1950-2013.nc"))
 
     da1 = stack_variables(ds1)
     da2 = stack_variables(ds2)
 
+    # FIXME: Thes test for variable order; use a membership test instead
     assert list(da1.multivar.values) == ["pr", "tasmax"]
     assert da1.multivar.attrs["_standard_name"] == [
         "precipitation_flux",