Skip to content

Restore support for non-contiguous PyArrays, fix f_contiguous, simplify copy(::PyArray)#623

Merged
stevengj merged 4 commits intoJuliaPy:masterfrom
JobJob:non-contig-pyarrays
Dec 14, 2018
Merged

Restore support for non-contiguous PyArrays, fix f_contiguous, simplify copy(::PyArray)#623
stevengj merged 4 commits intoJuliaPy:masterfrom
JobJob:non-contig-pyarrays

Conversation

@JobJob
Copy link
Copy Markdown
Contributor

@JobJob JobJob commented Dec 12, 2018

Changes by commit

  1. Add support for non-contiguous PyArrays: Restore support for non-contiguous PyArrays (I think they used to work before PyArray conversion speedups and PyArrayFromBuffer #487?)
  2. Fix f_contiguous for 1d arrays: without this non-contiguous 1d PyArrays could have the f_contig flag set incorrectly to true. This would cause an incorrect conversion to julia for such arrays with a convert(PyAny, non_contig_1d_array) (see the [1,0,1,0] example in Use native sizes by default when parsing buffer format string #621/the "1d non-contiguous" testset in this PR)
  3. Simplify Copy(::PyArray): With f_contiguous correct for 1d, copy(::PyArray) could be simplified

@JobJob JobJob force-pushed the non-contig-pyarrays branch from 6cbec28 to 8e5be56 Compare December 12, 2018 08:32
@JobJob
Copy link
Copy Markdown
Contributor Author

JobJob commented Dec 12, 2018

force-push

Naughty naughty...

@stevengj stevengj merged commit 093342b into JuliaPy:master Dec 14, 2018
@stevengj
Copy link
Copy Markdown
Member

Thanks!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants