[SciPy-user] problems numpy 0.9.4 and matplotlib

John Hunter jdhunter at ace.bsd.uchicago.edu
Mon Jan 23 09:53:43 CST 2006


>>>>> "Aurimas" == Aurimas Anskaitis <aanskis at yahoo.com> writes:

    Aurimas>  I have some problems with numpy-0.9.4 and matplotlib
    Aurimas> 0.86.1. When I try to launch very simple script like
    Aurimas> this: import pylab from scipy import * y = array([1, 2,
    Aurimas> 3]) pylab.plot(y) pylab.show()
 
    Aurimas>  I get a long error message ending with line "ValueError:
    Aurimas> arrays must have same number of dimensions". May somebody
    Aurimas> explain me what's happening? Everything works just fine
    Aurimas> using numpy 0.9.2.
  
You need to make sure that matplotlib knows about which array type you
are using by setting the "numerix" setting in you matplotlib rc file,
typically found in ~/.matplotlib/matplotlibrc but you can find the
definitive location by running your script with --verbose-helpful
which the rc file location as well as which array package you are
using.  You will want to make sure your numerix setting is numpy.  

FYI, 

  from pylab import *

together with 

  from scipy import *

is a recipe for confusion, since both packages import a lot of names
with a overlapping names.

Hope this helps,
JDH



More information about the SciPy-user mailing list