Fix benchmark CI (#9013)

* [skip-ci] Fix benchmark CI

* [skip-ci] reduce warnings

* Fix indexing benchmark

Author: dcherian

.github/workflows/benchmarks.yml

@@ -28,8 +28,11 @@ jobs:
           environment-name: xarray-tests
           cache-environment: true
           cache-environment-key: "${{runner.os}}-${{runner.arch}}-py${{env.PYTHON_VERSION}}-${{env.TODAY}}-${{hashFiles(env.CONDA_ENV_FILE)}}-benchmark"
+          # add "build" because of https://github.com/airspeed-velocity/asv/issues/1385
           create-args: >-
             asv
+            build
+            mamba
 
 
       - name: Run benchmarks
@@ -47,9 +50,6 @@ jobs:
           asv machine --yes
           echo "Baseline:  ${{ github.event.pull_request.base.sha }} (${{ github.event.pull_request.base.label }})"
           echo "Contender: ${GITHUB_SHA} (${{ github.event.pull_request.head.label }})"
-          # Use mamba for env creation
-          # export CONDA_EXE=$(which mamba)
-          export CONDA_EXE=$(which conda)
           # Run benchmarks for current commit against base
           ASV_OPTIONS="--split --show-stderr --factor $ASV_FACTOR"
           asv continuous $ASV_OPTIONS ${{ github.event.pull_request.base.sha }} ${GITHUB_SHA} \

asv_bench/asv.conf.json

@@ -29,7 +29,7 @@
     // If missing or the empty string, the tool will be automatically
     // determined by looking for tools on the PATH environment
     // variable.
-    "environment_type": "conda",
+    "environment_type": "mamba",
     "conda_channels": ["conda-forge"],
 
     // timeout in seconds for installing any dependencies in environment
@@ -41,7 +41,7 @@
 
     // The Pythons you'd like to test against.  If not provided, defaults
     // to the current version of Python used to run `asv`.
-    "pythons": ["3.10"],
+    "pythons": ["3.11"],
 
     // The matrix of dependencies to test.  Each key is the name of a
     // package (in PyPI) and the values are version numbers.  An empty
@@ -72,8 +72,12 @@
         "sparse": [""],
         "cftime": [""]
     },
-
-
+    // fix for bad builds
+    // https://github.com/airspeed-velocity/asv/issues/1389#issuecomment-2076131185
+    "build_command": [
+        "python -m build",
+        "python -mpip wheel --no-deps --no-build-isolation --no-index -w {build_cache_dir} {build_dir}"
+    ],
     // Combinations of libraries/python versions can be excluded/included
     // from the set to test. Each entry is a dictionary containing additional
     // key-value pairs to include/exclude.

asv_bench/benchmarks/groupby.py

@@ -68,6 +68,7 @@ def setup(self, *args, **kwargs):
         self.ds2d_mean = self.ds2d.groupby("b").mean().compute()
 
 
+# TODO: These don't work now because we are calling `.compute` explicitly.
 class GroupByPandasDataFrame(GroupBy):
     """Run groupby tests using pandas DataFrame."""
 
@@ -111,11 +112,11 @@ def setup(self, *args, **kwargs):
             {
                 "b": ("time", np.arange(365.0 * 24)),
             },
-            coords={"time": pd.date_range("2001-01-01", freq="H", periods=365 * 24)},
+            coords={"time": pd.date_range("2001-01-01", freq="h", periods=365 * 24)},
         )
         self.ds2d = self.ds1d.expand_dims(z=10)
-        self.ds1d_mean = self.ds1d.resample(time="48H").mean()
-        self.ds2d_mean = self.ds2d.resample(time="48H").mean()
+        self.ds1d_mean = self.ds1d.resample(time="48h").mean()
+        self.ds2d_mean = self.ds2d.resample(time="48h").mean()
 
     @parameterized(["ndim"], [(1, 2)])
     def time_init(self, ndim):
@@ -127,15 +128,15 @@ def time_init(self, ndim):
     def time_agg_small_num_groups(self, method, ndim, use_flox):
         ds = getattr(self, f"ds{ndim}d")
         with xr.set_options(use_flox=use_flox):
-            getattr(ds.resample(time="3M"), method)().compute()
+            getattr(ds.resample(time="3ME"), method)().compute()
 
     @parameterized(
         ["method", "ndim", "use_flox"], [("sum", "mean"), (1, 2), (True, False)]
     )
     def time_agg_large_num_groups(self, method, ndim, use_flox):
         ds = getattr(self, f"ds{ndim}d")
         with xr.set_options(use_flox=use_flox):
-            getattr(ds.resample(time="48H"), method)().compute()
+            getattr(ds.resample(time="48h"), method)().compute()
 
 
 class ResampleDask(Resample):
@@ -154,13 +155,13 @@ def setup(self, *args, **kwargs):
             },
             coords={
                 "time": xr.date_range(
-                    "2001-01-01", freq="H", periods=365 * 24, calendar="noleap"
+                    "2001-01-01", freq="h", periods=365 * 24, calendar="noleap"
                 )
             },
         )
         self.ds2d = self.ds1d.expand_dims(z=10)
-        self.ds1d_mean = self.ds1d.resample(time="48H").mean()
-        self.ds2d_mean = self.ds2d.resample(time="48H").mean()
+        self.ds1d_mean = self.ds1d.resample(time="48h").mean()
+        self.ds2d_mean = self.ds2d.resample(time="48h").mean()
 
 
 @parameterized(["use_cftime", "use_flox"], [[True, False], [True, False]])

asv_bench/benchmarks/indexing.py

@@ -19,6 +19,7 @@
 }
 
 basic_assignment_values = {
+    "1scalar": 0,
     "1slice": xr.DataArray(randn((3, ny), frac_nan=0.1), dims=["x", "y"]),
     "1slice-1scalar": xr.DataArray(randn(int(ny / 3) + 1, frac_nan=0.1), dims=["y"]),
     "2slicess-1scalar": xr.DataArray(