Speed up Dataset._construct_dataarray (#4744)

* Speed up Dataset._construct_dataarray * Preserve ordering
pydata · Jan 5, 2021 · 9fefefb · 9fefefb
1 parent 8731acc
commit 9fefefb
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 4 deletions.
diff --git a/doc/whats-new.rst b/doc/whats-new.rst
@@ -33,11 +33,11 @@ Breaking changes
 
 New Features
 ~~~~~~~~~~~~
-
+- Performance improvement when constructing DataArrays. Significantly speeds up repr for Datasets with large number of variables.
+  By `Deepak Cherian <https://github.com/dcherian>`_
 
 Bug fixes
 ~~~~~~~~~
-
 - :py:meth:`DataArray.resample` and :py:meth:`Dataset.resample` do not trigger computations anymore if :py:meth:`Dataset.weighted` or :py:meth:`DataArray.weighted` are applied (:issue:`4625`, :pull:`4668`). By `Julius Busecke <https://github.com/jbusecke>`_.
 - :py:func:`merge` with ``combine_attrs='override'`` makes a copy of the attrs (:issue:`4627`).
 - By default, when possible, xarray will now always use values of type ``int64`` when encoding

diff --git a/xarray/core/dataset.py b/xarray/core/dataset.py
@@ -1320,8 +1320,9 @@ def _construct_dataarray(self, name: Hashable) -> "DataArray":
         needed_dims = set(variable.dims)
 
         coords: Dict[Hashable, Variable] = {}
-        for k in self.coords:
-            if set(self.variables[k].dims) <= needed_dims:
+        # preserve ordering
+        for k in self._variables:
+            if k in self._coord_names and set(self.variables[k].dims) <= needed_dims:
                 coords[k] = self.variables[k]
 
         if self._indexes is None: