MPL/Pylab install woes (still)

Community matplotlib-users
1

After some progress in installing matplotlib (after solving the numpy problem, thanks), I’ve hit another obstacle. Again, perhaps someone will recognise the symptoms or suggest the next place to look. The technical set up: OSX 10.4, MPL 0.87.5, numeric, numarray and numpy 1.0b5 installed, intel MacBook.

The issue: matplotlib installs but when pylab is imported, I get: :

14:28:56: Debug: …/src/common/object.cpp(224): assert “sm_classTable->Get(m_className) == NULL” failed: class already in RTTI table - have you used IMPLEMENT_DYNAMIC_CLASS() twice (may be by linking some object module(s) twice)?

…/src/common/object.cpp(224): assert “sm_classTable->Get(m_className) == NULL” failed: class already in RTTI table - have you used IMPLEMENT_DYNAMIC_CLASS() twice (may be by linking some object module(s) twice)?

Trace/BPT trap

and then dumped to the CLI. sm_classTable appears to be defined in the wxagg backend.

When matplotlib is built, I get a bunch of warnings like::

/usr/bin/ld: for architecture ppc

/usr/bin/ld: warning fat file: /usr/local/lib/libgcc_s.10.4.dylib does not contain an architecture that matches the specified -arch flag: ppc (file ignored)

when linking various agg and _nc_transforms pieces, as well as::

src/_ns_backend_agg.cpp:41:1: warning: “M_PI” redefined

In file included from /usr/include/math.h:26,

from /Developer/SDKs/MacOSX10.4u.sdk/usr/include/c++/4.0.0/cmath:52,

although these don’t terminate the build. It would seem that something is getting architectures mixed up, but I can’t identify what. Things I’ve done - reinstalled Python (made sure it was universal), updated Developer tools, cleared & reinstalled MPL. Any ideas on what to try next?

p

···

–

Dr Paul-Michael Agapow, VieDigitale / Institute of Animal Health

pma@…1244… / paul-michael.agapow@…1245…

2

After some progress in installing matplotlib (after solving the numpy
problem, thanks), I've hit another obstacle. Again, perhaps someone will
recognise the symptoms or suggest the next place to look. The technical set
up: OSX 10.4, MPL 0.87.5, numeric, numarray and numpy 1.0b5 installed,
intel MacBook.

The issue: matplotlib installs but when pylab is imported, I get: :

    14:28:56: Debug: ../src/common/object.cpp(224): assert
"sm_classTable->Get(m_className) == NULL" failed: class
already in RTTI table - have you used IMPLEMENT_DYNAMIC_CLASS() twice (may
be by linking some object module(s) twice)?
    ../src/common/object.cpp(224): assert
"sm_classTable->Get(m_className) == NULL" failed: class
already in RTTI table - have you used IMPLEMENT_DYNAMIC_CLASS() twice (may
be by linking some object module(s) twice)?
    Trace/BPT trap

and then dumped to the CLI. ``sm_classTable`` appears to be defined in the
wxagg backend.

Do you have the universal wxpython installed? If so, make sure you
set WX_CONFIG to point to the correct wx-config executable.

When matplotlib is built, I get a bunch of warnings like::

   /usr/bin/ld: for architecture ppc
   /usr/bin/ld: warning fat file:
/usr/local/lib/libgcc_s.10.4.dylib does not contain an
architecture that matches the specified -arch flag: ppc (file ignored)

I am guessing you are on an intel machine. These can safely be ignored.

when linking various agg and _nc_transforms pieces, as well as::

   src/_ns_backend_agg.cpp:41:1: warning: "M_PI" redefined
   In file included from /usr/include/math.h:26,
      from
/Developer/SDKs/MacOSX10.4u.sdk/usr/include/c++/4.0.0/cmath:52,

This should be wrapped with an ifdef, but again they can be ignored for now.

although these don't terminate the build. It would seem that something is
getting architectures mixed up, but I can't identify what. Things I've done
- reinstalled Python (made sure it was universal), updated Developer tools,
cleared & reinstalled MPL. Any ideas on what to try next?

Sounds like wx is your problem right now, so please try my suggestion above.

- Charlie