- Update to 1.7
* Breaking: the skip_axes argument iter_indices() function now
applies the skipped axes before broadcasting, not after. This
behavior is more generally useful and matches how functions
with stacking work (e.g., np.cross or np.matmul). The best way
to get the old behavior is to broadcast the arrays/shapes
together first. The skip_axes in iter_indices must be either
all negative or all nonnegative to avoid ambiguity. A future
version may add support for specifying different skip axes for
each shape.
* iter_indices() no longer requires the skipped axes specified by
skip_axes to be broadcast compatible.
* New method isvalid() to check if an index is valid on a given
shape.
* New function broadcast_shapes() which is the same as
np.broadcast_shapes() except it also allows specifying a set of
skip_axes which will be ignored when broadcasting.
* New exceptions BroadcastError and AxisError which are used by
iter_indices() and broadcast_shapes().
* Fix some test failures with the latest version of NumPy.
* Fix some tests that didn’t work properly when run against the
* sdist.
* The sdist now includes relevant testing files.
- Drop patches fixed upstream
* ndindex-pr133-ragged.patch gh#Quansight-Labs/ndindex#133
* ndindex-pr147-numpy.patch gh#Quansight-Labs/ndindex#147
- Refresh custom-pytest.patch gh#Quansight-Labs/ndindex#150
- Add ndindex-pr159-py312.patch gh#Quansight-Labs/ndindex#159