将 Numpy 与 pypy 结合使用
我正在使用一些 numpy 工具（主要是数组），我想用 pypy 运行脚本，但我无法让它工作。
我收到的错误是：ImportError: No module named multiarray.
我检查了 multiarray.so 文件是否位于核心文件夹中。
有人可以告诉我，第一：是否可以做我想做的事情，第二：我该怎么做？
I am using some numpy tools (mainly arrays) and I wanted to run the script with pypy, but i can't make it work.
The error that i get is: ImportError: No module named multiarray.
I checked if the multiarray.so file was in the core folder.
Can someone tell me if first: is possible to do what I am trying to do and second: How can I do it?
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（5）
我刚刚发布了一篇博客文章解释了状态以及计划是什么。简而言之，numpy 不能与 PyPy 的 cpyext 一起使用，即使可以，使用起来也会太慢。
I've just posted a blog post explaining what's the status and what's the plan. In short numpy will not work with PyPy's cpyext and even if it does, it would be too slow for usage.
其他答案都很旧了。
这是
对“实现的功能”的完全不科学的测量numpypy状态页面pypy 博客中关于 numpy 的一些帖子：
The other answers are quite old.
Here is the
the completely unscientific measure of "implemented functions" onnumpypy status pageSome posts from the pypy blog about numpy:
Numpy 状态和构建指令最近已更改。 numpy 有一个特殊版本，已移植到 PyPy。如果您想获取最新说明，只需查看 PyPy 博客，获取有关 Numpy 的最新文章。在撰写本文时，最新说明位于这篇文章中，其中编译为：
对于已实现的内容和未实现的内容，您可以检查此页面： http://buildbot.pypy.org/ numpy-status/latest.html
Numpy status and build instruction has been changed recently. There is a special version of numpy which is ported to PyPy. If you want to get latest instruction just check PyPy blog for a latest article about Numpy. For the time of writing the latest instruction are in this post, which compiles to:
For what is implemented and what not you can check this page: http://buildbot.pypy.org/numpy-status/latest.html
以前的答案现在已经过时了:)
http://morepypy.blogspot.com/2011/08/pypy-16-kickass- panda.html
CPython 扩展模块 API 已得到改进，现在支持更多扩展。有关支持哪一种的信息，请参阅我们的兼容性 wiki。
对 NumPy 的初步支持：此版本包括与 PyPy JIT 集成的非常快的 NumPy 模块的预览。
Previous answers now are obsolete :)
http://morepypy.blogspot.com/2011/08/pypy-16-kickass-panda.html
The CPython extension module API has been improved and now supports many more extensions. For information on which one are supported, please refer to our compatibility wiki.
Preliminary support for NumPy: this release includes a preview of a very fast NumPy module integrated with the PyPy JIT.
据我所知，Numpy 从未成功移植到 PyPy，而且这样做并非易事，因为 PyPy 的 C API 与 cPython 不同。
另请参阅http://ademan.wordpress.com/category/code/pypy/对于尝试过的人的博客。
As far as I know, Numpy has never been succesfully ported to PyPy, and it is not trivial to do so as the C API for PyPy is different from the cPython one.
See also http://ademan.wordpress.com/category/code/pypy/ for the blog of someone that tried.