Begin forwarded message:

Date: July 19, 2006 4:12:46 PM EDT
Cc: Jon Guyer <[EMAIL PROTECTED]>
Subject: Re: Getting FiPy working on Windows XP

Hi Daniel and Jon,
   Thanks for your feedback.  I have had some progress on executing the
single examples.  Although I had initially used the IDLE GUI, I recently
had tried other ways.  But returning to the
IDLE GUI has worked.  I saw "rainbow" color plots when executing (F5 key)
for both
      'C:\Python23\FiPy-1.1
      \examples\diffusion\steadyState\mesh50x50\input.py', and
      'C:\Python23\FiPy-1.1\examples\diffusion\mesh20x20.py'       (here I
      get a prompt on the Python shell after the first rainbow plot, and
      hitting ENTER produces a second plot).
    Trying execution of Jon's suggested example 'C:\Python23\FiPy-1.1
\examples\diffusion\mesh1D.py' produced no plot and this error message:
      Traceback (most recent call last):
        File "C:\Python23\FiPy-1.1\examples\diffusion\mesh1D.py", line 779,
      in ?
          exec(fipy.tests.doctestPlus._getScript())
        File "<string>", line 38, in ?
        File "C:\Python23\Lib\site-packages\fipy\viewers\__init__.py", line
      88, in make
          raise ImportError, "Failed to import a viewer: %s" % str(errors)
      ImportError: Failed to import a viewer: ['gistViewer: cannot import
      name GistViewer', 'gnuplotViewer: cannot import name GnuplotViewer',
      'matplotlibViewer: cannot import name MatplotlibViewer',
      'mayaviViewer: Mayavi does not display useful 1D data', "multiViewer:
      'module' object has no attribute 'make'", "tsvViewer: 'module' object
      has no attribute 'make'", "viewer: 'module' object has no attribute
      'make'"]

     Thanks for your help!  Now I feel like I can proceed to code my
problem and give it a try.
      It's OK with me if you post my email inquiry to users.
Terry McDaniel
Seagate Research     (209) 295-6735



             Daniel Wheeler                                            
             <[EMAIL PROTECTED]                                         
             ist.gov>                                                   To
             No Phone Info             [EMAIL PROTECTED]    
             Available                                                  cc
                                       Jon Guyer <[EMAIL PROTECTED]>      
                                                                   Subject
             07/19/2006 12:27          Re: Getting FiPy working on Windows
             PM                        XP                              










Hi Terry,

Thanks for your interest. Would you mind if I forward this email and the
reply to
our mailing list so that it is archived?

Hopefully, FiPy will help you with your work and we are happy to
provide assistance where
we can.

As far as the tests in the examples directory are concerned, you probably
just need to set
the PYTHONPATH to point at 'C:\Python23\FiPy-1.1' then the tests should be
able to
see the required modules.

You also mentioned that when you run the examples you are not seeing any
plots on the screen.
What happens when you open

    C:\Python23\examples\diffusion\steadyState\mesh50x50\input.py

in IDLE and then run the module (press F5)? Is it that nothing shows up at
all or do you get an error message?

If it is that nothing shows up at all, try the following command at the
IDLE command line (assuming you have Matplotlib
installed):

      >>> import pylab
      >>> pylab.plot((1,2,3))
      >>> pylab.show()

Do you get something on the screen?

Cheers

Daniel Wheeler

On Jul 19, 2006, at 4:24 PM, [EMAIL PROTECTED] wrote:

      Gentlemen,
          I'm a physicist trying to solve the heat conduction PDE in (z,t)
      for an
      application we have.  The problem has a time-dependent heat source
      and the
      possibility of a temperature-dependent thermal
      conductivity in one film in a stack.  Of course, I have abrupt
      discontinuities between the different films in the stack (at least in
      the
      model).  To date, I have run into seemingly insurmountable
      difficulties in
      getting toward a solution in Mathcad, Mathematica, and MATLAB.
      Then I
      came across FiPy on the web, and it looks more capable, so I was
      encouraged.  However, after attempting the Windows installation, and
      running some of the tests, I cannot seem to get past a string of
      errors,
      and I have not yet seen any plot of a result in any example.   Here
      is a
      recent log of the errors from running 'test.py' in
      'C:\Python23\FiPy-1.1
      \examples'   (See attached file: Python_test_errors3.doc) .  I can't
      really
      decipher it, but I suspect that FiPy is having some trouble locating
      files.
      Here is a screen shot of the main Python directory on my C drive
      where all
      this stuff is placed   (See attached file: Python23_directory.gif) .
          I'm not necessarily a computer specialist - I'm a modeler.  I
      tried to
      follow the order and instructions for installation from the site
      'path' in the environmental variables was a little vague to me, so
      all that
      I did manually was set the Windows XP environmental variable / system
      variable PYTHONPATH to 'C:\Python23'.
      I notice that the installation
      procedure apparently added automatically a string
      'C:\Python23;C:\Python23\Scripts;C:\Python23\Enthought\MingW\bin;C:\Python23\Enthought\SWIG-1.3.24'
       to the Path system variable.  Perhaps my environment is still not
      completely set up.
          I am eager to get on to solving a PDE if I can get around the
      installation hiccoughs.  Thanks in advance for any suggestions you
      might
      offer.  I am reachable at the phone below, or of course at this email
      address.   I might be able to link up on Webex or NetMeeting if
      desired.
      Terry McDaniel
      Seagate Research     (209) 295-6735
      <Python_test_errors3.doc>
      <Python23_directory.gif>


Daniel Wheeler







Daniel Wheeler



Reply via email to