matplotlib-users Mailing List for matplotlib

2010/11/1 Timothy W. Hilton <hi...@me...>:
> [...]
>
> I want to have blue water, some other (bright) color for missing data,
> and a nice-looking color transition (matplotlib.cm.Blues or something
> similar) for the valid data over land (values from 0 to 50).  The
> Cookbook example at
> <http://www.scipy.org/Cookbook/Matplotlib/Plotting_Images_with_Special_Values>
> addresses my problem, but I cannot get it to work.  After changing
> instances of matplotlib.numerix to numpy, I get a long list of
> exceptions, the last of which is
> TypeError: __call__() got an unexpected keyword argument 'bytes'.
> This has to do with sentinelNorm, I think, but I'm not sure how to fix it.

I think I would tackle this by writing a Norm that doesn't change
negative values, and you might mask then the oceans by -0.5, and the
nans by -1.5.  Then, you might create a colormap comprised of the
ocean color for [-1, 0] and the nan color for [-2, -1], and for the
normal normed range [0, 1] the normal Blues cm.  Have a look at cm.py
and _cm.py how it works.  Basically, you can specify for all sections
of the colormap the left and right color.  So you can mix discrete
maps with continuous ones, because the continuous ones are just
linearly interpolated with matching colors for left/right at each
boundary position.  Looking at the code will clarify things a lot I
believe.

I don't know what went wrong with the cb example you said.  From a
quick look, it seems to have "sentinel rgb values", but this is not
what we want, right?

> Eventually I would like to sub-classify missing data by the type of
> missing input that caused a missing value, but for now a single
> missing data color is enough.

That would be possible with the approach above, by just adding
sections below zero.

_cm.py: Definitions of colormaps, like Blues.
cm.py: among other things, how to load such specifications.
colors.py: Defines Colormaps, and Norms.  Have a look at both of them,
esp. at :class:`Normalize`.  I would subclass the Norm mentioned above
from Normalize.

I hope this helps you,
Friedrich

Hi,

A.S.: Ralf, wasn't the reason why we shifted to Vincent's build
machine running 10.5, because numpy didn't run on 10.4 when built on
10.6?

2010/11/2 pablomos <pa...@pr...>:
> I attempted to install Pylab on my own computer and so far have had no
> success. I followed these steps:
> 1) downloaded Python 2.6 for OS 10.4
> 2) downloaded the corresponding numpy and installed it (successfully)

> 3) downloaded the corresponding matplotlib
> (matplotlib-1.0.0-python.org-py2.6-macosx10.4.dmg) and installed it [One
> possible cause for the error is that, though the package includes
> "macosx10.4", the actual installing package (which one can see if one tries
> a customized install) says "macosx10.5".]

This is due to a bug in bdist_mpkg probably, used to generate those
binaries on a 10.5 system, although the MACOSX_DEPLOYMENT_TARGET was
10.4 or 10.3 (as I strongly believe), the installer outputs somehow
"-10.5" when run.  I don't know details.

Of course it is contradictory to my above assumtion that the build was
carried out on a 10.6 ...

> 4) Attempted to run Python and import Pylab. I got an error related to
> _tkinter not being found (even though it is installed). I looked all over
> the forums for this error, and it seems that the only solution is to edit
> the setup.py file and recompile matplotlib. However, because I installed
> from a binary, I cannot do this. So I attempted to install from source:

Can you give more details on the _tkinter error, is it matplotlib
specific, or does it mean the Python _tkinter if such one exists at
all?  It's kinda strange.

> 5) I downloaded matplotlib-1.0.0.tar.gz and untarred it. Then I ran:
> 'make' (which supposedly takes care of finding out where all the Mac stuff
> is located) and

For Python packages, ``make`` is *usually* not used.  Indeed,
matplotlib is an exception here (see below).

> 'python setup.py build', the default command for building Python modules.
> Now I get a ton of errors, the first one being:
> freetype2: found, but unknown version (no pkg-config)
> * WARNING: Could not find 'freetype2' headers in any
> * of '.', './freetype2'.

You got it.  If you have freetype2 and libpng installed on your
system, just have a look at setupext.py, and add the installation root
path (most likely '/usr/local') to the 'darwin' section.  It is left
out because of problems on > 10.4, but on your 10.4 system it should
work.  It's the most easy solution if.

If you don't have freetype2 installed / libpng, and you really want to
compile from source (if we cannot sort out the _tkinter thingy), then
there are two ways: Hard-core installation by compiling them yourself,
or assisted by make.osx.  I personally cannot give you advice with the
more easy make.osx way, but be aware of that also this compiles and
installs the libs.  For the hard-core way you might find my docs on
that useful (it's intended for 10.6, so feedback from 10.4 users is
welcome): http://vincentdavis.info/Shared/Docs/matplotlib-installation/build/html/.
 Vincent is hosting that site, gratefully acknowledged.

> As a third try, I attempted to install using easy_install, a tool for
> installing Python modules:
> 6) I ran 'easy_install matplotlib'. This runs very quickly (suspicious)
> 7) The same error as in step 4 comes up when trying to import pylab
>  I went into
> /Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/
> and deleted all matplotlib-related stuff, thinking that maybe easy_install
> didn't do anything because matplotlib was already installed

> 9) Repeated step 6, and saw the same errors that came up when trying to
> install from source

You mean easy_install wanted to install from source??

Btw, the binaries *should* work.  It looks like kind of failure of the
binaries.  Might be that I'm missing something, though.

Hope you find this useful,
Friedrich

[Don't forget to "reply to all" and not only me.]