Enable pandas type checking (pydata#9213)

* remove pandas from ignore missing imports

* add Any to dim arg of concat as placeholder

* allow sequence of np.ndarrays as coords in dataArray constructor

* fix several typing issues in tests

* fix more types

* more fixes

* more typing...

* we are getting there?

* who might have guessed it... more typing

* continue fixing typing issues

* fix some typed_ops

* fix last non-typed-ops errors

* update typed ops

* remove useless DaskArray type in scalar or array type

* fix missing import in type_checking

* fix import

* improve cftime offsets typing

* fix classvars

* fix some checks

* fix a broken test

* improve typing of test_concat

* fix broken concat

* add whats-new

---------

Co-authored-by: Maximilian Roos <[email protected]>

Author: headtr1ck

doc/whats-new.rst

@@ -82,6 +82,8 @@ Documentation
 Internal Changes
 ~~~~~~~~~~~~~~~~
 
+- Enable typing checks of pandas (:pull:`9213`).
+  By `Michael Niklas <https://github.com/headtr1ck>`_.
 
 .. _whats-new.2024.06.0:
 

pyproject.toml

@@ -120,7 +120,6 @@ module = [
   "netCDF4.*",
   "netcdftime.*",
   "opt_einsum.*",
-  "pandas.*",
   "pint.*",
   "pooch.*",
   "pyarrow.*",