matplotlib-users Mailing List for matplotlib

> .....
> But you can also create your own color maps at any time using the
> cm.get_cmap function
>
> >>> jet512 = cm.get_cmap('jet', 512)
> >> imshow(X, cmap=jet512)
>
> Should work.
>
Yes, thanks.
Just one more question. Recently (in the last couple of versions I 
think), there has been a change and now the 'o' line markers (circles) 
are filled in by default.
For example:

    plot(xp,yp,'ok')

gives sold black circles.

I would like my circles to not be filled; just have the border so that 
the inside is transparent.
Is this something that is settable?

Thanks,

-- 
Peter Groszkowski Gemini Observatory
Tel: +1 808 974-2509 670 N. A'ohoku Place
Fax: +1 808 935-9235 Hilo, Hawai'i 96720, USA

>>>>> "Vineet" == Vineet Jain <vi...@al...> writes:

    Vineet> After making changes to has_data I get the following:
    Vineet> (0.0, 400.0) (0.0, 400.0) (0.0, 400.0)

    Vineet> which is not correct it should be:

    Vineet> (300.0, 400.0) (100.0, 400.0) (100.0, 400.0)

I'm not sure this is incorrect.  matplotlib distinguishes between the
data limits and the view limits.  The latter are automatically
adjusted to include the data limits but may incorporate more.  For
example, the x data limits in [0.1, 10] may produce view limits
[0,10].

If you need the view limits to always equal the data limits, you could
provide a custom ticker, as described in
http://matplotlib.sourceforge.net/matplotlib.ticker.html and
illustrated in the example
http://matplotlib.sourceforge.net/examples/custom_ticker1.py.

You might want to verify that the data limits are correct by printing

  ax.dataLim.intervalx().get_bounds()  # the x limits
  ax.dataLim.intervaly().get_bounds()  # the y limits

where ax is your Axes instance.  You can compare these with

  ax.viewLim.intervalx().get_bounds()  # the x limits
  ax.viewLim.intervaly().get_bounds()  # the y limits

JDH

>>>>> "Kenneth" == Kenneth McDonald <ken...@sb...> writes:

    Kenneth> Does any exist? I looked through the distribution and
    Kenneth> didn't see anything, but then, that doesn't prove
    Kenneth> anything :-)

There is no official documentation, but there are a few resources

  * The class documentation at
    http://matplotlib.sourceforge.net/classdocs.html

  * the examples embedding_in*.py in the examples directory.

  * examples/pythonic_matplotlib.py describes the translation from the
    matlab style interface to the pythonic, OO interface

  * Another good resource is matlab.py.  This is just a thin wrapper
    to the API, so you can look and see how it is done there,
    translating all the gcf() calls to your figure instance, all the
    gca() calls to your axes instance, and so on.

Other than that, you can post here...

I'm working on some additional documentation but it is not ready yet.

JDH

>>>>> "Mathias" == Mathias Franzius <mat...@we...> writes:

    Mathias> Dear NG, I have a problem mwith the installation of
    Mathias> matplotlib under redhat 9 and python 2.3.2. I first
    Mathias> updated all relevant packages (see [1]).  Building
    Mathias> matplotlib with default setup.py seemd to be ok, except
    Mathias> two types of warnings (see [2]). Install showed no errors
    Mathias> as well.  When importing matplotlib I get the following
    Mathias> error:

The warnings you posted are completely harmless.  

    Mathias> What could be the problem?  The directory
    Mathias> /usr/lib/python2.3/site-packages/matplotlib/ contains
    Mathias> transforms.py and _transforms.so but no _transforms.py -
    Mathias> is that ok?

Yes, this is OK.  There is no _transforms.py

My guess is you are trying to run matplotlib from the build dir.  You
cannot do this, because python finds the matplotlib src dir named
matplotlib and tries to load that.  There is no _transforms module in
that dir, but there is in site-packages so you should be OK.  cd into
another directory, ie your home dir or the examples dir, and try from
there.

This problem bites a number of users.  Perhaps we should rename the
base matplotlib python dir in the src tree to something like
matplotlibpy.

JDH

>>>>> "Peter" == Peter Groszkowski <pgr...@ge...> writes:

    Peter> Hi: How many colors do images created via:

    Peter>         imshow(Zi, cmap=cm.jet)

    Peter> (Zi = some data matrix) have?

    Peter> Are these true color? 256? Is there a simple way to define
    Peter> these things?

The following image parameters can be configured from your rc file

### images
image.aspect : free          # free | preserve
image.interpolation  : bilinear  # see help(imshow) for options
image.cmap   : jet           # gray | jet
image.lut    : 256           # the size of the colormap lookup table
image.origin : upper         # lower | upper


The image.lut parameter controls the size of the lookup table.  You
can change the default in the rc file, or dynamically in a single
python session using the rc function.  Eg,

 # default cmap is now 100 level grayscale by cm.jet and cm.gray unaffected
 >>> rc('image', lut=100, cmap='gray')
 >>> imshow(X)  # show X with default cmap


But you can also create your own color maps at any time using the
cm.get_cmap function

  >>> jet512 = cm.get_cmap('jet', 512)
  >> imshow(X, cmap=jet512)

Should work.

JDH

Dear NG,

I have a problem mwith the installation of matplotlib under redhat 9 and 
python 2.3.2. I first updated all relevant packages (see [1]).
Building matplotlib with default setup.py seemd to be ok, except two 
types of warnings (see [2]). Install showed no errors as well.
When importing matplotlib I get the following error:
 >>> from matplotlib import matlab
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "matplotlib/matlab.py", line 142, in ?
     from axes import Axes
   File "matplotlib/axes.py", line 9, in ?
     from artist import Artist
   File "matplotlib/artist.py", line 4, in ?
     from transforms import identity_transform
   File "matplotlib/transforms.py", line 180, in ?
     from _transforms import Value, Point, Bbox, Affine
ImportError: No module named _transforms

What could be the problem?
The directory /usr/lib/python2.3/site-packages/matplotlib/
contains transforms.py and _transforms.so but no _transforms.py - is 
that ok?

Thanks for any help,
    Mathias

--------
[1]: zlib, zlib-devel, libpng, libpng-devel, freetype, freetype-devel, 
freetype-utils, gtk2-devel, gtk+-devel, pygtk2, glib-devel, 
pygtk2-devel, gnome-libs-devel, pygtk2-libglade,
tcl, tk, tkinter

[2]: In file included from /usr/include/python2.3/Python.h:8,
                  from CXX/Objects.hxx:9,
                  from CXX/Extensions.hxx:18,
                  from src/_transforms.h:10,
                  from src/_transforms.cpp:2:
/usr/include/python2.3/pyconfig.h:847:1: warning: "_POSIX_C_SOURCE" 
redefined
In file included from 
/usr/include/c++/3.2.2/i386-redhat-linux/bits/os_defines.h:39,
                  from 
/usr/include/c++/3.2.2/i386-redhat-linux/bits/c++config.h:34,
                  from /usr/include/c++/3.2.2/functional:53,
                  from src/_transforms.cpp:1:
/usr/include/features.h:131:1: warning: this is the location of the 
previous definition