MAINT: Give a more helpful error for bad axis specifications.

This specifically addresses passing things like axis=[0,1] which
gave an error message that an int was required.  The real error
was that if the axis isn't None or a tuple, it was simply wraped
in a tuple, so the ufunc that is eventually called ended up with
a list.  The error message is matches that in
core/src/multiarray/conversion_utils.c::PyArray_ConvertMultiAxis.

1 files changed, 4 insertions, 0 deletions

diff --git a/numpy/linalg/linalg.py b/numpy/linalg/linalg.py
index e8f7a8ab1..e35c9ac97 100644
--- a/numpy/linalg/linalg.py
+++ b/numpy/linalg/linalg.py
@@ -2074,6 +2074,10 @@ def norm(x, ord=None, axis=None, keepdims=False):
     if axis is None:
         axis = tuple(range(nd))
     elif not isinstance(axis, tuple):
+        try:
+            axis = int(axis)
+        except:
+            raise TypeError("'axis' must be None, an integer or a tuple of integers")
         axis = (axis,)
 
     if len(axis) == 1: