matplotlib-users Mailing List for matplotlib

Is there a way to figure out the optimal dpi for imshow when saving as a png
so that all the pixels are visible?

I do something like the following:

import matplotlib.pyplot as plt
plt.clf()
image = np.random.normal(0,1, (1000,1000))
plt.imshow(I, cmap=plt.cm.gray, interpolation='nearest', aspect='equal')
plt.axis('off')
plt.title('title')
plt.savefig('~/tmp/test.png, dpi=600)

If the size of 'image' is too large or dpi is too small, the elements in the
matrix get blended. I would like to set the dpi to the minimal necessary so
that the elements are distinct.

Any tips appreciated.

Aha, I did not have VC++ 2008 EE on my PC, so it actually didn't compile 
anything.
Now that I have it, I get a dependency problem with Freetype2. The build 
command fails with:
c:\python26\matpotlib\src\ft2font.h(13) : fatal error C1083: Cannot open 
include
 file: 'ft2build.h': No such file or directory

I downloaded the windows binaries of freetype2 in gnuwin32 project, but 
it doesn't seem to help. Or do I need to download freetype2 source and 
compile that. If so, maybe there is a detailed (step-by-step) guide 
installing the dependecies (libpng, zlib, freetype), because they don't 
seem to have windows installers like numpy has that actually works with 
matplotlib.

Any helpl appreciated.
When might we see matplotlib binaries for python 2.6?

Andrew Straw wrote:
> Romi Agar wrote:
>   
>> Hi!
>>
>> I'm having a bit difficulty getting matplotlib to run under windows 
>> (vista x64) with python 2.6.
>> I downloaded the source from svn, ran the build and install commands, 
>>     
>
> Does that mean "python setup.py install"?
>
>   
>> then copied the content of /build/lib.win32-2.6
>> to Lib/site-packages folder. 
>>     
>
> Hmm, why did you have to do that? Doesn't the install step do that
> automatically? Or did it break?
>
> But when I try to import pylab I get the
>   
>> following error message:
>>
>> Traceback (most recent call last):
>>   File "<pyshell#0>", line 1, in <module>
>>     import pylab
>>   File "C:\Python26\lib\site-packages\pylab.py", line 1, in <module>
>>     from matplotlib.pylab import *
>>   File "C:\Python26\lib\site-packages\matplotlib\pylab.py", line 207, in 
>> <module>
>>     from matplotlib import mpl  # pulls in most modules
>>   File "C:\Python26\lib\site-packages\matplotlib\mpl.py", line 1, in 
>> <module>
>>     from matplotlib import artist
>>   File "C:\Python26\lib\site-packages\matplotlib\artist.py", line 5, in 
>> <module>
>>     from transforms import Bbox, IdentityTransform, TransformedBbox, 
>> TransformedPath
>>   File "C:\Python26\lib\site-packages\matplotlib\transforms.py", line 
>> 34, in <module>
>>     from matplotlib._path import affine_transform
>> ImportError: No module named _path
>>
>> So, where might I find the _path module be?
>>     
>
> This gets compiled from C sources. Do you a C compiler on your system?
> "Python 2.6 is built with MicroSoft Visual Studio 2008 (aka VC++ v 9.0),
> but the free VC++ 2008 Express Edition compiles Python extensions just
> fine", according to http://mien.sourceforge.net/docs/platform_win.html#py26
>
>
>   
>> Thanks in advance,
>> Romi
>>
>> ------------------------------------------------------------------------------
>> Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
>> powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
>> easily build your RIAs with Flex Builder, the Eclipse(TM)based development
>> software that enables intelligent coding and step-through debugging.
>> Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>     
>
>   

On Thu, Mar 19, 2009 at 2:33 PM, Eric Firing <ef...@ha...> wrote:
>
>
> def noclip(ax):
>     "Turn off all clipping in axes ax; call immediately before drawing"
>     ax.set_clip_on(False)
>     artists = []



Or even better::

for o in fig.findobj():
    o.set_clip_on(False)

findobj is an artist method that recursive searches all the artists
contained in it -- you can optionally specify the type of artist you want
returned.  See


http://matplotlib.sourceforge.net/examples/pylab_examples/findobj_demo.html

JDH

Hatch, Sara J wrote:
> Matplotlib Folks,
> 
>  
> 
> How do I turn off all clipping when making a plot?  It seems like 
> everything has a set_clip_on argument, but I couldn’t figure out how to 
> set all of these to False without explicitly doing so in every plot 
> call. I would assume that there is a way to do this using an rcParam or 
> the matplotlibrc file?
> 

No there isn't, and I think this is the first time this question has 
come up.  Usually some clipping is desired.  I doubt the need for 
absolutely no clipping is common enough to justify an rcParam entry.  If 
you think there is a common use case that should be supported, though, 
please elaborate.

Every artist has a set_clip_on() method, so to turn off all clipping you 
are stuck having to find all the artists and turn off clipping on each 
individually.  Maybe something like this (untested):

def noclip(ax):
     "Turn off all clipping in axes ax; call immediately before drawing"
     ax.set_clip_on(False)
     artists = []
     artists.extend(ax.collections)
     artists.extend(ax.patches)
     artists.extend(ax.lines)
     artists.extend(ax.texts)
     artists.extend(ax.artists)
     for a in artists:
         a.set_clip_on(False)

I suspect this will not necessarily take care of everything; there may 
be compound artists that do not define their own set_clip_on method to 
propagate down to the sub-artists.

Eric

Romi Agar wrote:
> Hi!
> 
> I'm having a bit difficulty getting matplotlib to run under windows 
> (vista x64) with python 2.6.
> I downloaded the source from svn, ran the build and install commands, 

Does that mean "python setup.py install"?

> then copied the content of /build/lib.win32-2.6
> to Lib/site-packages folder. 

Hmm, why did you have to do that? Doesn't the install step do that
automatically? Or did it break?

But when I try to import pylab I get the
> following error message:
> 
> Traceback (most recent call last):
>   File "<pyshell#0>", line 1, in <module>
>     import pylab
>   File "C:\Python26\lib\site-packages\pylab.py", line 1, in <module>
>     from matplotlib.pylab import *
>   File "C:\Python26\lib\site-packages\matplotlib\pylab.py", line 207, in 
> <module>
>     from matplotlib import mpl  # pulls in most modules
>   File "C:\Python26\lib\site-packages\matplotlib\mpl.py", line 1, in 
> <module>
>     from matplotlib import artist
>   File "C:\Python26\lib\site-packages\matplotlib\artist.py", line 5, in 
> <module>
>     from transforms import Bbox, IdentityTransform, TransformedBbox, 
> TransformedPath
>   File "C:\Python26\lib\site-packages\matplotlib\transforms.py", line 
> 34, in <module>
>     from matplotlib._path import affine_transform
> ImportError: No module named _path
> 
> So, where might I find the _path module be?

This gets compiled from C sources. Do you a C compiler on your system?
"Python 2.6 is built with MicroSoft Visual Studio 2008 (aka VC++ v 9.0),
but the free VC++ 2008 Express Edition compiles Python extensions just
fine", according to http://mien.sourceforge.net/docs/platform_win.html#py26


> 
> Thanks in advance,
> Romi
> 
> ------------------------------------------------------------------------------
> Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
> powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
> easily build your RIAs with Flex Builder, the Eclipse(TM)based development
> software that enables intelligent coding and step-through debugging.
> Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

The easiest solution is the "title" keyword for subplot:

pylab.subplot(221, title='A Title for 221')

I'm using matplotlib/pylab 0.98.3

-- NTroutman


Tommy Grav wrote:
> 
> I have a plot that is divided into four subplots.
> 
> pylab.figure()
> pylab.subplot(221)
> pylab.plot(a,b,"k-")
> pylab.subplot(222)
> pylab.plot(a,b,"k-")
> pylab.subplot(223)
> pylab.plot(a,b,"k-")
> pylab.subplot(224)
> pylab.plot(a,b,"k-")
> 
> I would like to add a title to the entire plot, but pylab.title() only
> applies to the most recent subplot. I have tried
> 
> pylab.figure()
> pylab.subplot(111)
> pylab.title("Title Here")
> pylab.subplot(221)
> pylab.plot(a,b,"k-")
> pylab.subplot(222)
> pylab.plot(a,b,"k-")
> pylab.subplot(223)
> pylab.plot(a,b,"k-")
> pylab.subplot(224)
> pylab.plot(a,b,"k-")
> 
> but this does not work as I do not create a plot for
> the subplot(111) instance. Is there some way of getting
> the type of title I want easily?
> 
> Cheers
>    Tommy
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
> 

-- 
View this message in context: http://www.nabble.com/title-when-using-subplot-tp15066004p22605721.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

Hi!

I'm having a bit difficulty getting matplotlib to run under windows 
(vista x64) with python 2.6.
I downloaded the source from svn, ran the build and install commands, 
then copied the content of /build/lib.win32-2.6
to Lib/site-packages folder. But when I try to import pylab I get the 
following error message:

Traceback (most recent call last):
  File "<pyshell#0>", line 1, in <module>
    import pylab
  File "C:\Python26\lib\site-packages\pylab.py", line 1, in <module>
    from matplotlib.pylab import *
  File "C:\Python26\lib\site-packages\matplotlib\pylab.py", line 207, in 
<module>
    from matplotlib import mpl  # pulls in most modules
  File "C:\Python26\lib\site-packages\matplotlib\mpl.py", line 1, in 
<module>
    from matplotlib import artist
  File "C:\Python26\lib\site-packages\matplotlib\artist.py", line 5, in 
<module>
    from transforms import Bbox, IdentityTransform, TransformedBbox, 
TransformedPath
  File "C:\Python26\lib\site-packages\matplotlib\transforms.py", line 
34, in <module>
    from matplotlib._path import affine_transform
ImportError: No module named _path

So, where might I find the _path module be?

Thanks in advance,
Romi

Awesome.  Thanks for the quick fix and the tip about using 'rm'.

Ryan

On Thu, Mar 19, 2009 at 9:43 AM, Michael Droettboom <md...@st...> wrote:

> They're actually different fonts.  The one on the left is Vera Sans, the
> one on right is Stix Sans.  There is a bug where regular (meaning Vera) text
> in math was getting mapping to Stix Sans when mathtext.fontset is set to
> stixsans.  (This should now be fixed in SVN).
>
> All that said, fixing this "bug" will mean it's now fairly easy to get an
> unintended mix of Vera Sans and Stix Sans in math expressions.  For example,
> "$A \bf{A}$" will draw the first A in Vera and the second in STIX. There's
> no easy way around that, since we don't currently support rich text for
> non-math fonts.  To avoid this mix of fonts within the $ $, you would
> probably want to set mathtext.default to "rm".
>
> Mike
>
> Ryan May wrote:
>
>> Mike,
>>
>> Is there supposed to be a difference in the size of text printed with and
>> without mathtext?  Here's a simple script that shows what I'm talking about:
>>
>> import matplotlib.pyplot as plt
>> f = plt.figure()
>> ax1 = f.add_subplot(1,2,1)
>> ax1.set_title('Hello')
>> ax2 = f.add_subplot(1,2,2)
>> ax2.set_title('$Hello$')
>> plt.show()
>>
>> And here's my matplotlibrc:
>>
>> backend      : GtkAgg
>> mathtext.fontset : stixsans
>> font.size           : 10.0
>> savefig.dpi       : 100      # figure dots per inch
>> mathtext.default : regular
>>
>> I've attached a copy of what that looks like on my system.
>>
>> Sorry I keep finding these things. :)
>> Ryan
>>
>> --
>> Ryan May
>> Graduate Research Assistant
>> School of Meteorology
>> University of Oklahoma
>> Sent from: Norman Oklahoma United States.
>> ------------------------------------------------------------------------
>>
>> ------------------------------------------------------------------------
>>
>>
>> ------------------------------------------------------------------------------
>> Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
>> powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
>> easily build your RIAs with Flex Builder, the Eclipse(TM)based development
>> software that enables intelligent coding and step-through debugging.
>> Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>
>>
>
> --
> Michael Droettboom
> Science Software Branch
> Operations and Engineering Division
> Space Telescope Science Institute
> Operated by AURA for NASA
>
>


-- 
Ryan May
Graduate Research Assistant
School of Meteorology
University of Oklahoma
Sent from: Norman Oklahoma United States.

They're actually different fonts.  The one on the left is Vera Sans, the 
one on right is Stix Sans.  There is a bug where regular (meaning Vera) 
text in math was getting mapping to Stix Sans when mathtext.fontset is 
set to stixsans.  (This should now be fixed in SVN).

All that said, fixing this "bug" will mean it's now fairly easy to get 
an unintended mix of Vera Sans and Stix Sans in math expressions.  For 
example, "$A \bf{A}$" will draw the first A in Vera and the second in 
STIX. There's no easy way around that, since we don't currently support 
rich text for non-math fonts.  To avoid this mix of fonts within the $ 
$, you would probably want to set mathtext.default to "rm".

Mike

Ryan May wrote:
> Mike,
>
> Is there supposed to be a difference in the size of text printed with 
> and without mathtext?  Here's a simple script that shows what I'm 
> talking about:
>
> import matplotlib.pyplot as plt
> f = plt.figure()
> ax1 = f.add_subplot(1,2,1)
> ax1.set_title('Hello')
> ax2 = f.add_subplot(1,2,2)
> ax2.set_title('$Hello$')
> plt.show()
>
> And here's my matplotlibrc:
>
> backend      : GtkAgg
> mathtext.fontset : stixsans
> font.size           : 10.0
> savefig.dpi       : 100      # figure dots per inch
> mathtext.default : regular
>
> I've attached a copy of what that looks like on my system.
>
> Sorry I keep finding these things. :) 
>
> Ryan
>
> -- 
> Ryan May
> Graduate Research Assistant
> School of Meteorology
> University of Oklahoma
> Sent from: Norman Oklahoma United States.
> ------------------------------------------------------------------------
>
> ------------------------------------------------------------------------
>
> ------------------------------------------------------------------------------
> Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
> powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
> easily build your RIAs with Flex Builder, the Eclipse(TM)based development
> software that enables intelligent coding and step-through debugging.
> Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
> ------------------------------------------------------------------------
>
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>   

-- 
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA

That fixed it for me, thanks.

Ryan

On Thu, Mar 19, 2009 at 7:54 AM, Michael Droettboom <md...@st...> wrote:

> Thanks.  I think I have this fixed in SVN again...
>
> Basically what's happening is that Freetype's hinting imposes a vertical
> scale on the glyph that wasn't being taken into account.
>
> Mike
>
> Ryan May wrote:
>
>> Mike,
>>
>> I've found another case of the funk baseline:
>>
>> import matplotlib.pyplot as plt
>> plt.title('$Z_{DR}$ (from ts)')
>> plt.show()
>>
>> I've attached an image of what I see. Here's my matplotlibrc:
>>
>> backend      : GtkAgg
>> mathtext.fontset : stixsans
>> mathtext.default : regular
>> font.size           : 10.0
>> savefig.dpi       : 100
>>
>> Ryan
>>
>> --
>> Ryan May
>> Graduate Research Assistant
>> School of Meteorology
>> University of Oklahoma
>> Sent from: Norman Oklahoma United States.
>> ------------------------------------------------------------------------
>>
>> ------------------------------------------------------------------------
>>
>>
>> ------------------------------------------------------------------------------
>> Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
>> powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
>> easily build your RIAs with Flex Builder, the Eclipse(TM)based development
>> software that enables intelligent coding and step-through debugging.
>> Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>
>>
>
> --
> Michael Droettboom
> Science Software Branch
> Operations and Engineering Division
> Space Telescope Science Institute
> Operated by AURA for NASA
>
>


-- 
Ryan May
Graduate Research Assistant
School of Meteorology
University of Oklahoma
Sent from: Norman Oklahoma United States.

Thanks.  I think I have this fixed in SVN again...

Basically what's happening is that Freetype's hinting imposes a vertical 
scale on the glyph that wasn't being taken into account.

Mike

Ryan May wrote:
> Mike,
>
> I've found another case of the funk baseline:
>
> import matplotlib.pyplot as plt
> plt.title('$Z_{DR}$ (from ts)')
> plt.show()
>
> I've attached an image of what I see. Here's my matplotlibrc:
>
> backend      : GtkAgg
> mathtext.fontset : stixsans
> mathtext.default : regular
> font.size           : 10.0
> savefig.dpi       : 100
>
> Ryan
>
> -- 
> Ryan May
> Graduate Research Assistant
> School of Meteorology
> University of Oklahoma
> Sent from: Norman Oklahoma United States.
> ------------------------------------------------------------------------
>
> ------------------------------------------------------------------------
>
> ------------------------------------------------------------------------------
> Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
> powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
> easily build your RIAs with Flex Builder, the Eclipse(TM)based development
> software that enables intelligent coding and step-through debugging.
> Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
> ------------------------------------------------------------------------
>
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>   

-- 
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA

cneff wrote:
> Hi guys,
> 
> I've tried to google this and look through the examples but its not quite
> working for me. Say I have two sets of data I want to make contour plots out
> of
> 
> from pylab import *
> 
> x=arange(-3.0,3.0,.025)
> 
> y=arange(-2.0,2.0,.025)
> 
> X,Y = meshgrid(x,y)
> 
> Z1 = mlab.bivariate_normal(X, Y, 1.0, 1.0, 0.0, 0.0)
> Z2 = mlab.bivariate_normal(X, Y, 1.5, 0.5, 1, 1)
> 
> Now if I would go
> 
> plt1 = subplot(211)
> contourf(X,Y,Z1)
> colorbar()
> 
> plt2 = subplot(212)
> contourf(X,Y,Z2)
> colorbar()
> 
> we would see that the same colors correspond to different numerical values,
> because the ranges of Z1 and Z2 are different.  I want it to be defined on
> the same range, so that red on plt1 corresponds to the same numerical Z
> value as red on plt2.  How do I go about doing that?

Instead of relying on autoscaling to set the color levels, set them 
explicitly to the same set of values in both calls to contourf by adding 
a fourth argument.

e.g.

levs = arange(0,1.01,0.1)

...

contourf(X, Y, Z1, levs)

...

contourf(X, Y, Z2, levs)

...

Eric

Hi guys,

I've tried to google this and look through the examples but its not quite
working for me. Say I have two sets of data I want to make contour plots out
of

from pylab import *

x=arange(-3.0,3.0,.025)

y=arange(-2.0,2.0,.025)

X,Y = meshgrid(x,y)

Z1 = mlab.bivariate_normal(X, Y, 1.0, 1.0, 0.0, 0.0)
Z2 = mlab.bivariate_normal(X, Y, 1.5, 0.5, 1, 1)

Now if I would go

plt1 = subplot(211)
contourf(X,Y,Z1)
colorbar()

plt2 = subplot(212)
contourf(X,Y,Z2)
colorbar()

we would see that the same colors correspond to different numerical values,
because the ranges of Z1 and Z2 are different.  I want it to be defined on
the same range, so that red on plt1 corresponds to the same numerical Z
value as red on plt2.  How do I go about doing that?
-- 
View this message in context: http://www.nabble.com/Using-the-same-color-range-for-multiple-plots-tp22592487p22592487.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

Matplotlib Folks,

How do I turn off all clipping when making a plot?  It seems like everything has a set_clip_on argument, but I couldn't figure out how to set all of these to False without explicitly doing so in every plot call. I would assume that there is a way to do this using an rcParam or the matplotlibrc file?

Any help would be appreciated!

Thanks,
Sara

-------------------------------------------------------------------
Sara Jean Hatch
Inner Planet Mission Analysis Group
Guidance, Navigation, & Control Section
NASA - Jet Propulsion Laboratory
4800 Oak Grove Drive
M/S: 301-150
Pasadena, CA 91109
Phone: (818) 354-8723
-------------------------------------------------------------------

crwe crwe wrote:
> Looking at the implementation, it's really simple and i will do
> without matplotlib altogether and roll my own. 

It sounds like you have a solution, but if you need to test a LOT of 
points, it can be efficient to rasterize, and then get a lightning fast 
check.

It that case, though, I'd be inclined to use something like PIL's 
ImageDraw, rather than MPL -- all of MPL's scaling, anti-aliasing, etc. 
will just make it harder.

-Chris



-- 
Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/OR&R            (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chr...@no...