ENH: copy ndarrays if they are encountered in setArrayFromSequence - #3696

almost solves the performance issue Graeme Belt reported on the list. (3.4s -> 0.55s)
http://mail.scipy.org/pipermail/numpy-discussion/2013-September/067607.html

the rest is inefficient short circuit reduction for multidimensional arrays.

Looks good to me
On 7 Sep 2013 12:24, "Julian Taylor" [email protected] wrote:

By avoiding the elementwise copy array([a,b,c]) is now as

fast as vstack([a,b,c])

You can merge this Pull Request by running

git pull https://github.com/juliantaylor/numpy copy-array-create

Or view, comment on, or merge it at:

#3696
Commit Summary

• ENH: copy ndarrays if they are encountered in setArrayFromSequence

File Changes

• M numpy/core/src/multiarray/ctors.chttps://github.com/ENH: copy ndarrays if they are encountered in setArrayFromSequence #3696/files#diff-0(74)
• M numpy/core/tests/test_multiarray.pyhttps://github.com/ENH: copy ndarrays if they are encountered in setArrayFromSequence #3696/files#diff-1(42)

Patch Links:

• https://github.com/numpy/numpy/pull/3696.patch
• https://github.com/numpy/numpy/pull/3696.diff