Improve behaviour of DataTree accessor (#32)

* better default for datatree group handling

* add test for datatree accessor group behaviour

* lint

.pre-commit-config.yaml

@@ -1,6 +1,6 @@
 repos:
 - repo: https://github.com/pre-commit/pre-commit-hooks
-  rev: v4.6.0
+  rev: v5.0.0
   hooks:
   - id: check-added-large-files
   - id: check-toml
@@ -11,7 +11,7 @@ repos:
   - id: trailing-whitespace
 
 - repo: https://github.com/astral-sh/ruff-pre-commit
-  rev: v0.4.9
+  rev: v0.6.9
   hooks:
     - id: ruff
       args: [ --fix, --exit-non-zero-on-fix ]

src/arviz_stats/accessors.py

@@ -1,6 +1,7 @@
 """ArviZ stats accessors."""
 
 import warnings
+from collections.abc import Hashable
 
 import numpy as np
 import xarray as xr
@@ -214,27 +215,40 @@ def _process_input(self, group, method, allow_non_matching=True):
             )
             return self._obj
         raise ValueError(
-            f"Group {group} not available in DataTree. Present groups are {self._obj.children}"
+            f"Group {group} not available in DataTree. Present groups are {self._obj.children} "
+            f"and the DataTree itself is named {self._obs.name}"
         )
 
     def _apply(self, fun_name, group, **kwargs):
-        allow_non_matching = False
-        if isinstance(group, str):
+        hashable_group = False
+        if isinstance(group, Hashable):
             group = [group]
-            allow_non_matching = True
-        return DataTree.from_dict(
+            hashable_group = True
+        out_dt = DataTree.from_dict(
             {
                 group_i: xr.Dataset(
                     {
                         var_name: get_function(fun_name)(da, **update_kwargs_with_dims(da, kwargs))
                         for var_name, da in self._process_input(
-                            group_i, fun_name, allow_non_matching=allow_non_matching
+                            # if group is a single str/hashable that doesn't match the group
+                            # name, still allow it and apply the function to the top level of
+                            # the provided datatree
+                            group_i,
+                            fun_name,
+                            allow_non_matching=hashable_group,
                         ).items()
                     }
                 )
                 for group_i in group
             }
         )
+        if hashable_group:
+            # if group was a string/hashable, return a datatree with a single node
+            # (from the provided group) as the root of the DataTree
+            return out_dt[group[0]]
+        # if group was a sequence, return a DataTree with multiple groups in the 1st level,
+        # as many groups as requested
+        return out_dt
 
     def filter_vars(self, group="posterior", var_names=None, filter_vars=None):
         """Access and filter variables of the provided group."""

tests/base/test_diagnostics.py

@@ -7,6 +7,7 @@
 import pandas as pd
 import pytest
 from arviz_base import load_arviz_data, xarray_var_iter
+
 from arviz_stats.base import array_stats
 
 # For tests only, recommended value should be closer to 1.01-1.05

tests/base/test_stats_utils.py

@@ -4,11 +4,12 @@
 #  pylint: disable=no-member,unnecessary-lambda-assignment
 import numpy as np
 import pytest
-from arviz_stats.base.stats_utils import logsumexp as _logsumexp
-from arviz_stats.base.stats_utils import make_ufunc, not_valid
 from numpy.testing import assert_array_almost_equal
 from scipy.special import logsumexp
 
+from arviz_stats.base.stats_utils import logsumexp as _logsumexp
+from arviz_stats.base.stats_utils import make_ufunc, not_valid
+
 
 @pytest.mark.parametrize("ary_dtype", [np.float64, np.float32, np.int32, np.int64])
 @pytest.mark.parametrize("axis", [None, 0, 1, (-2, -1)])

tests/test_accessors.py

@@ -0,0 +1,48 @@
+# pylint: disable=redefined-outer-name
+"""Test accessors.
+
+Accessor methods are very short, with the bulk of the computation/processing
+handled by private methods. Testing this shared infrastructural methods
+is the main goal of this module even if it does so via specific "regular" methods.
+"""
+
+import numpy as np
+import pytest
+from arviz_base import from_dict
+from datatree import DataTree
+
+
+@pytest.fixture(scope="module")
+def idata():
+    return from_dict(
+        {
+            "posterior": {
+                "a": np.random.normal(size=(4, 100)),
+                "b": np.random.normal(size=(4, 100, 3)),
+            },
+            "posterior_predictive": {
+                "y": np.random.normal(size=(4, 100, 7)),
+            },
+        }
+    )
+
+
+def test_accessors_available(idata):
+    assert hasattr(idata, "azstats")
+    assert hasattr(idata.posterior.ds, "azstats")
+    assert hasattr(idata.posterior["a"], "azstats")
+
+
+def test_datatree_single_group(idata):
+    out = idata.azstats.ess(group="posterior")
+    assert isinstance(out, DataTree)
+    assert not out.children
+    assert out.name == "posterior"
+
+
+def test_datatree_multiple_groups(idata):
+    out = idata.azstats.ess(group=["posterior", "posterior_predictive"])
+    assert isinstance(out, DataTree)
+    assert len(out.children) == 2
+    assert "posterior" in out.children
+    assert "posterior_predictive" in out.children

tests/test_psense.py

@@ -1,10 +1,11 @@
 import os
 
 from arviz_base import convert_to_datatree
-from arviz_stats import psense, psense_summary
 from numpy import isclose
 from numpy.testing import assert_almost_equal
 
+from arviz_stats import psense, psense_summary
+
 file_path = os.path.join(os.path.dirname(__file__), "univariate_normal.nc")
 uni_dt = convert_to_datatree(file_path)
 

tests/test_utils.py

@@ -5,6 +5,7 @@
 import numpy as np
 import pytest
 from arviz_base import from_dict, rcParams
+
 from arviz_stats.base.dataarray import dataarray_stats
 from arviz_stats.utils import ELPDData, get_function, get_log_likelihood
 

tox.ini

@@ -9,9 +9,10 @@ isolated_build_env = build
 
 [gh-actions]
 python =
-    3.10: check, py310
+    3.10: py310
     3.11: py311
-    3.12: py312
+    3.12: py312, check
+    3.13: py313
 
 [testenv]
 basepython =