In file included from root_numpy/src/_librootnumpy.cpp:460:
In file included from /Users/DC/Downloads/build/include/TClass.h:25:
In file included from /Users/DC/Downloads/build/include/TDictionary.h:45:
In file included from /Users/DC/Downloads/build/include/TNamed.h:26:
In file included from /Users/DC/Downloads/build/include/TObject.h:18:
/Users/DC/Downloads/build/include/Rtypes.h:37:10: fatal error: 'atomic' file not found
#include <atomic>
^
1 warning and 1 error generated.
error: command 'gcc' failed with exit status 1
I already updated my gcc. Is there any way for me to get that mysterious atomic file and put it in it’s place?
Configured with: --prefix=/Library/Developer/CommandLineTools/usr --with-gxx-include-dir=/usr/include/c++/4.2.1
Apple LLVM version 7.0.2 (clang-700.1.81)
Target: x86_64-apple-darwin14.5.0
Thread model: posix
But I am not sure if that is what has been used to build numpy? I have installed anaconda/jupyter with python2 along with different packages before I got this latest version. Any suggestions?
But I figured I had to uninstall and reinstall anaconda to rebuild numpy I guess and I sort of screwed things up more as the notebook can’t find the gcc file. So I can’t work with root_numpy in the root --notebook.
Ah macos, that’s a relevant piece of information I missed… You should not use gcc if you built or downloaded the default configuration of ROOT on macos. Instead you should teach numpy to build with clang.
If you install gcc from whatever package manager you also need to build your own ROOT with it, tell python to use that etc. It’s a mess…
Perhaps I have a very basic question now, how do you build numpy with clang exactly? My idea is to build from source again but I am not sure how to teach it to build with clang.