rewritten checking uniqueness of multi-index level names

Author: Benoit Bovy

xarray/core/dataarray.py

@@ -19,7 +19,8 @@
 from .dataset import Dataset
 from .pycompat import iteritems, basestring, OrderedDict, zip
 from .variable import (as_variable, Variable, as_compatible_data, Coordinate,
-                       default_index_coordinate)
+                       default_index_coordinate,
+                       assert_unique_multiindex_level_names)
 from .formatting import format_item
 
 
@@ -81,14 +82,7 @@ def _infer_coords_and_dims(shape, coords, dims):
                                  'length %s on the data but length %s on '
                                  'coordinate %r' % (d, sizes[d], s, k))
 
-        if v.ndim == 1:
-            idx_level_names = v.to_coord().level_names or []
-            for n in idx_level_names:
-                if n in level_names:
-                    raise ValueError('found duplicate MultiIndex level '
-                                     'name %r for coordinates %r and %r'
-                                     % (n, k, level_names[n]))
-                level_names[n] = k
+    assert_unique_multiindex_level_names(new_coords)
 
     return new_coords, dims
 

xarray/core/dataset.py

@@ -224,7 +224,6 @@ def __init__(self, data_vars=None, coords=None, attrs=None,
             coords = {}
         if data_vars is not None or coords is not None:
             self._set_init_vars_and_dims(data_vars, coords, compat)
-            self._check_multiindex_level_names()
         if attrs is not None:
             self.attrs = attrs
         self._initialized = True
@@ -251,21 +250,6 @@ def _set_init_vars_and_dims(self, data_vars, coords, compat):
         self._coord_names = coord_names
         self._dims = dims
 
-    def _check_multiindex_level_names(self):
-        """Check for uniqueness of MultiIndex level names
-        """
-        level_names = {}
-        for c in self._coord_names:
-            v = self._variables[c]
-            if v.ndim == 1 and v._in_memory:
-                idx_level_names = v.to_coord().level_names or []
-                for n in idx_level_names:
-                    if n in level_names:
-                        raise ValueError('found duplicate MultiIndex level '
-                                         'name %r for coordinates %r and %r'
-                                         % (n, c, level_names[n]))
-                    level_names[n] = c
-
     @classmethod
     def load_store(cls, store, decoder=None):
         """Create a new dataset from the contents of a backends.*DataStore

xarray/core/merge.py

@@ -2,7 +2,8 @@
 
 from .alignment import deep_align
 from .utils import Frozen
-from .variable import as_variable, default_index_coordinate
+from .variable import (as_variable, default_index_coordinate,
+                       assert_unique_multiindex_level_names)
 from .pycompat import (basestring, OrderedDict)
 
 
@@ -110,7 +111,7 @@ def merge_variables(
         If provided, variables are always taken from this dict in preference to
         the input variable dictionaries, without checking for conflicts.
     compat : {'identical', 'equals', 'broadcast_equals', 'minimal'}, optional
-        Type of equality check to use wben checking for conflicts.
+        Type of equality check to use when checking for conflicts.
 
     Returns
     -------
@@ -278,6 +279,7 @@ def merge_coords_only(objs, priority_vars=None):
     """
     expanded = expand_variable_dicts(objs)
     variables = merge_variables(expanded, priority_vars)
+    assert_unique_multiindex_level_names(variables)
     return variables
 
 
@@ -319,6 +321,7 @@ def align_and_merge_coords(objs, compat='minimal', join='outer',
     expanded = expand_variable_dicts(aligned)
     priority_vars = _get_priority_vars(aligned, priority_arg, compat=compat)
     variables = merge_variables(expanded, priority_vars, compat=compat)
+    assert_unique_multiindex_level_names(variables)
 
     return variables
 
@@ -380,6 +383,7 @@ def merge_core(objs, compat='broadcast_equals', join='outer', priority_arg=None,
 
     priority_vars = _get_priority_vars(aligned, priority_arg, compat=compat)
     variables = merge_variables(expanded, priority_vars, compat=compat)
+    assert_unique_multiindex_level_names(variables)
 
     dims = calculate_dimensions(variables)
 

xarray/core/variable.py

@@ -1,4 +1,5 @@
 from datetime import timedelta
+from collections import defaultdict
 import functools
 import itertools
 import warnings
@@ -1297,3 +1298,28 @@ def concat(variables, dim='concat_dim', positions=None, shortcut=False):
         return Coordinate.concat(variables, dim, positions, shortcut)
     else:
         return Variable.concat(variables, dim, positions, shortcut)
+
+
+def assert_unique_multiindex_level_names(variables):
+    """Check for uniqueness of MultiIndex level names in all given
+    variables.
+
+    Not public API. Used for checking consistency of DataArray and Dataset
+    objects.
+    """
+    level_names = defaultdict(list)
+    for var_name, var in variables.items():
+        if isinstance(var._data, PandasIndexAdapter):
+            idx_level_names = var.to_coord().level_names
+            if idx_level_names is not None:
+                for n in idx_level_names:
+                    level_names[n].append(var_name)
+
+    duplicate_level_names = {k: v for k, v in level_names.items()
+                             if len(v) > 1}
+    if duplicate_level_names:
+        duplicate_str = '\n'.join(['level %r found in %s'
+                                   % (k, ' and '.join(v))
+                                   for k, v in duplicate_level_names.items()])
+        raise ValueError('conflicting MultiIndex level names:\n%s'
+                         % duplicate_str)