We need some more information.
Can you set verbose.level to debug-annoying in your
~/.matplotlib/matplotlibrc file?
Also, if you have the development tools (XCode) installed on your mac,
can you run python in gdb?
gdb python
GNU gdb Red Hat Linux (6.3.0.0-1.153.el4_6.2rh)
Copyright 2004
It looks like matplotlib was built for a different version of iconv than
what you have on your machine. This may be a scisoft packaging problem,
you may want to bring this question to their attention. Perhaps the
version of OSX that scisoft was built on/for is different than what
you're
Is there anything I can do to update it by myself in the meanwhile?
Thanks,
Nino
Michael Droettboom wrote:
It looks like matplotlib was built for a different version of iconv than
what you have on your machine. This may be a scisoft packaging problem,
you may want to bring this question to
You could try rebuilding matplotlib yourself. There are some OS-X
directions here:
http://www.astro.washington.edu/owen/BuildingMatplotlibForMac.html
This may mean, however, that it won't play nice with other parts of
scisoft. But this really is a matter of how and where scisoft was built
Hi Mike,
I re-install the astronomy scisoft package and now the segmentation
fault is not anymore present when I import matplotlib.mathtext.
But when I import pylab I have the following error.
I think it's something related with my backend choise. I tried to change
all the GUI possibilities in
Thanks,
I just had to replace the /usr/lib/libiconv.2.dylib with the one I have
in /sw/lib/ directory .
Now the GTK, GTKAgg backend work.
Nino
Michael Droettboom wrote:
You could try rebuilding matplotlib yourself. There are some OS-X
directions here:
Great. You may still want to let the scisoft guys know about this so
others won't hit up against it.
Cheers,
Mike
Antonino Cucchiara wrote:
Thanks,
I just had to replace the /usr/lib/libiconv.2.dylib with the one I
have in /sw/lib/ directory .
Now the GTK, GTKAgg backend work.
Nino
