0

I've googled this for a long time to no avail, I hope someone can point out a fix.

I've installed both Python 2.7 and 3.4 along with various packages for Python 2.7 such as matplotlib and numpy. I'm trying to install matplotlib and its dependencies into Python 3.4. I'm on Windows 7 (64 bit), the numpy package comes from http://www.lfd.uci.edu/~gohlke/pythonlibs/#numpy

When I right-click to install the numpy py3.4 version, I get the message "Python 3.4 required which was not found in the registry". How do I proceed? Thanks very much.

user3885774
  • 75
  • 1
  • 1
  • 5
  • 1
    did you install the 32 or 64 bit package? – Padraic Cunningham Jul 28 '14 at 21:50
  • @PadraicCunningham 64 bit – user3885774 Jul 28 '14 at 21:55
  • 1
    You should use the anaconda distribution, unless you have a really good reason to install manually. – mdurant Jul 28 '14 at 22:00
  • How did you install python 3? is it 64 bit? – Padraic Cunningham Jul 28 '14 at 22:03
  • @mdurant I see, I'd read that scientific package installation had dependency complications, that PIP wasn't able to handle, so it was necessary to not use PIP. – user3885774 Jul 28 '14 at 22:16
  • @PadraicCunningham I just tried to find the Python 3.4 MSI and found that it's not the 64 bit version. I'm going to get the 64 bit version, it should fix the problem. Thanks. – user3885774 Jul 28 '14 at 22:21
  • I am unaware of any such complications – mdurant Jul 28 '14 at 22:22
  • no worries, that is almost certainly the issue – Padraic Cunningham Jul 28 '14 at 22:25
  • @mdurant I was flipping between many kinds of documentation, so I might've read it wrong (the caveats possibly directed at developers rather than users), anyway, this is a link where I read about the dependency complications:https://packaging.python.org/en/latest/platforms.html#installing-scientific-packages – user3885774 Jul 28 '14 at 22:29
  • I believe that's about the difficulty of using (ana)conda from within an existing python installation, rather than using the anaconda installation itself, which comes with many useful libraries such as numpy already installed, and makes switching between python platforms easily. – mdurant Jul 28 '14 at 22:58

0 Answers0