matplotlib-users Mailing List for matplotlib

In article <CA65B53A.23941%wcc...@lb...>,
 Bill Carithers <wcc...@lb...> wrote:

> Hi all,
> 
> I had matplotlib-0.99.1.1 working with Python 2.6 on OS X v10.6(Snow
> Leopard). Today I upgraded to Lion (v10.7) and matplotlib had disappeared.
> Lion ships with Apple¹s version of Python 2.7. Even when I look back at
> Python 2.6 in the /System/Library/Frameworks/Python/Version/Python2.6, it
> seems to be gone. OK, I thought I would install matplotlib-1.0.1 in Python
> 2.7 using easy_install. It failed when unable to find the headers for numpy
> (terminal output below). Lion comes with numpy 1.5.1 pre-installed. I tried
> re-installing numpy but that didn¹t help so now I¹m stuck.
> 
> Any suggestions?

I personally would not try to mess with Apple's python. It's safer to 
leave it alone, and it is not upgrade very often.

My personal favorite option:
- install python.org's python
- install the official binary installers for numpy and matplotlib (both 
of which assume python.org's python)

You can use other standard distributions of Python including:
- fink
- MacPorts
- Enthought's python distribution
most or all of these come with their own versions of numpy and 
matplotlib (though in the case of fink and MacPorts you have to 
explicitly install them).

-- Russell

The navigation toolbar is ugly ;-)
I would like to change the icons, let's say for the 'home' position.
How can I do that?
--

E

> -----Original Message-----
> From: Marston Johnston [mailto:she...@gm...]
> Sent: Thursday, April 05, 2012 9:11 AM
> To: mat...@li...
> Subject: Re: [Matplotlib-users] 3D layered plot
> 
> Hi Eric,
> 
> For some reason I cannot see your post on the webpage but I get your
> emails.
> My python doesn't have: art3d.poly_collection_2d_to_3d()
> Is this an old function that has been deprecated? I'm using matplotlib
> version 1.1.0
> 
> /M
> 

I'm also running matplotlib version 1.1.0.  Perhaps someone else can comment, but as far as I know poly_collection_2d_to_3d is part of version 1.1.0.  How does it fail when you try to run the example I gave? Can you import mpl_toolkits.mplot3d.art3d? 

Hello,

This function 
<http://matplotlib.sourceforge.net/api/pyplot_api.html#matplotlib.pyplot.plotfile>, 
matplotlib.pyplot.plotfile, seems cannot plot from an ASCII table from 
stdin. Is there anyway to add this feature?

In addition, is there any command line interface for this function 
available?

Thanks!

Hong

Hi Eric,

For some reason I cannot see your post on the webpage but I get your emails.
My python doesn't have: art3d.poly_collection_2d_to_3d()
Is this an old function that has been deprecated? I'm using matplotlib version 1.1.0

/M

On Apr 5, 2012, at 10:47 AM, Marston wrote:

> 
> This plot is possible in MATLAB but I would like to do this in matplotlib.
> See attached fig.
> In matlab the code is:
> 
> a=peaks(20);
> b=peaks(20)+2*rand(20);
> c=rand(20);
> figure; hold on;
> ha=pcolor(a);
> hb=pcolor(b);
> hc=pcolor(c);
> set(hb,'zdata',0*b+5)
> set(hc,'zdata',0*c+10)
> 
> This plot can then be rotated, as is done in the figure, to what I want to
> accomplish in matplotlib.
> http://old.nabble.com/file/p33568798/set_matlab_image.jpg
> set_matlab_image.jpg 
> 
> Is this possible, if so, how?
> 
> -- 
> View this message in context: http://old.nabble.com/3D-layered-plot-tp33568798p33568798.html
> Sent from the matplotlib - users mailing list archive at Nabble.com.
> 
> 
> ------------------------------------------------------------------------------
> Better than sec? Nothing is better than sec when it comes to
> monitoring Big Data applications. Try Boundary one-second 
> resolution app monitoring today. Free.
> http://p.sf.net/sfu/Boundary-dev2dev
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

It sort of is:

import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
import mpl_toolkits.mplot3d.art3d as art3d import numpy as np

x, y = np.mgrid[0:2*np.pi:16j, 0:2*np.pi:16j]

C = np.cos(2*x[:-1,:-1]) + np.sin(y[:-1,:-1])

fig = plt.figure()
ax = fig.add_subplot(111, projection='3d') p = ax.pcolor(x, y, C) art3d.poly_collection_2d_to_3d(p, 0)

q = ax.pcolor(x, y, C)
art3d.poly_collection_2d_to_3d(q, 5)

o = ax.pcolor(x, y, C)
art3d.poly_collection_2d_to_3d(o, 10)


ax.set_xlim([x[0,0], x[-1,0]])
ax.set_ylim([y[0,0], y[0,-1]])
ax.set_zlim([0,10])


plt.show()


-Eric

> -----Original Message-----
> From: Marston [mailto:she...@gm...]
> Sent: Thursday, April 05, 2012 4:48 AM
> To: mat...@li...
> Subject: [Matplotlib-users] 3D layered plot
> 
> 
> This plot is possible in MATLAB but I would like to do this in
> matplotlib.
> See attached fig.
> In matlab the code is:
> 
> a=peaks(20);
> b=peaks(20)+2*rand(20);
> c=rand(20);
> figure; hold on;
> ha=pcolor(a);
> hb=pcolor(b);
> hc=pcolor(c);
> set(hb,'zdata',0*b+5)
> set(hc,'zdata',0*c+10)
> 
> This plot can then be rotated, as is done in the figure, to what I want
> to
> accomplish in matplotlib.
> http://old.nabble.com/file/p33568798/set_matlab_image.jpg
> set_matlab_image.jpg
> 
> Is this possible, if so, how?
> 
> --
> View this message in context: http://old.nabble.com/3D-layered-plot-
> tp33568798p33568798.html
> Sent from the matplotlib - users mailing list archive at Nabble.com.
> 
> 
> -----------------------------------------------------------------------
> -------
> Better than sec? Nothing is better than sec when it comes to
> monitoring Big Data applications. Try Boundary one-second
> resolution app monitoring today. Free.
> http://p.sf.net/sfu/Boundary-dev2dev
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

This plot is possible in MATLAB but I would like to do this in matplotlib.
See attached fig.
In matlab the code is:

a=peaks(20);
b=peaks(20)+2*rand(20);
c=rand(20);
figure; hold on;
ha=pcolor(a);
hb=pcolor(b);
hc=pcolor(c);
set(hb,'zdata',0*b+5)
set(hc,'zdata',0*c+10)

This plot can then be rotated, as is done in the figure, to what I want to
accomplish in matplotlib.
http://old.nabble.com/file/p33568798/set_matlab_image.jpg
set_matlab_image.jpg 

Is this possible, if so, how?

-- 
View this message in context: http://old.nabble.com/3D-layered-plot-tp33568798p33568798.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

I solved this problem using pyplot and the following code:
        x = np.arange(22)                                                                                                                                                                                                             
        y = x                                                                                                                                                                                                                         
        window = 1                                                                                                                                                                                                                    
        tsize = 8                                                                                                                                                                                                                     
        plt.clf()                                                                                                                                                                                                                     
        plt.cla()   
        nrow = 0                                                                                                                                                                                                                      
        for k in np.sort(dlist.keys()): # 2d data arrays of size (x,y)                                                                                                                                                                                              
            if not doplot[k]: continue                                                                                                                                                                                                
            vdic = dlist[k]                                                                                                                                                                                                           
            for v in np.sort(vdic.keys()):                                                                                                                                                                                            
                plt.subplot(self.rows,self.cols,window)                                                                                                                                                                               
                if vlist[k] == 'RR':                                                                                                                                                                                                  
                    plt.title(v,fontsize=tsize)                                                                                                                                                                                       
                    plt.subplots_adjust(bottom=0.26,left=0.18,  # These
values are taken from the GUI                                                                                                                                                                      
                                       right=.90, top=0.94,             #
window after I adjusted the fig to my liking                                                                                                                                                            
                                       wspace=0.07,hspace=0.94)                                                                                                                                                                       
                cs = plt.contourf(x,y,vdic[v],levels[k],cmap=                                                                                                                                                                         
                                  cm.get_cmap('jet',len(levels[k])-1))                                                                                                                                                                
                plt.axis('off')                                                                                                                                                                                                       
                window += 1                                                                                                                                                                                                           
                                                                                                                                                                                                                                      
            ax = plt.gca()                                                                                                                                                                                                            
            pos = ax.get_position()                                                                                                                                                                                                   
            l,b,w,h = pos.bounds                                                                                                                                                                                                      
            print l,b,w,h                                                                                                                                                                                                             
            cax = plt.axes([l-0.668,b-0.02,0.719,0.01]) # setup colorbar
axes                                                                                                                                                         
            cbar = plt.colorbar(cs,cax=cax,orientation='horizontal')                                                                                                                                                                  
            cl = plt.getp(cbar.ax, 'xticklabels')                                                                                                                                                                                     
            plt.setp(cl, fontsize=8)                                                                                                                                                                                                  
            fx = 0.07                                                                                                                                                                                                                 
            fy = 0.9                                                                                                                                                                                                                  
            dy = 0.15                                                                                                                                                                                                                 
            yp = fy - (dy*nrow)                                                                                                                                                                                                       
            plt.figtext(fx,yp,units[k],fontsize=tsize+1)                                                                                                                                                                              
            nrow += 1                                                                                                                                                                                                                 
            #pdb.set_trace() # debugging python -m file.py                                                                                                                                                                            
        cmd = '%s%s%s' % ('ece3_',win,'_L91.pdf')                                                                                                                                                                                     
        plt.savefig(cmd,dpi=600,orientation='landscape')  

I also attached an example.


Marston wrote:
> 
> Hi,
> 
> I'm trying to create a plot, with subplots where each row of x plots have
> a common colorbar beneath it. Only the the top row will have titles. I've
> tried creating a function to do this but I only achieve partial success.
> Here is an image created in another program that I want to duplicate:
> 
>  http://old.nabble.com/file/p33544950/fig.jpg fig.jpg 
>  
> I made several functions to do this. Here's one of them:
> 
> def Plot(self,title,plist):
>    for k in np.sort(self.vdic.keys()):                                                  
>       plt.subplot(5,13,self.window)                                                        
>       cs = plt.contourf(22,22,np.squeeze(self.vdic[k]),plist['levels'],              
>                            cmap=cm.get_cmap('jet',len(plist['levels'])-1))                                        
>       plt.axis('off')                                                             
>       if title: plt.title(k,fontsize=tsize)                                       
>       window += 1                                                            
>   cbar = plt.colorbar(cax=plist['cax'],orientation='h')                           
>                     
> I've given up in getting the text on the left because every attempt using
> text fails.
> Now if this is a horrible way and you have a better idea, please feel free
> to share.
> I'm new at matplotlib and this is a great way to see how things can be
> done better
> and in different ways. 
> 

-- 
View this message in context: http://old.nabble.com/Subplot-array-and-colorbar-tp33544950p33568733.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

On Wed, Mar 28, 2012 at 2:27 PM, Jouni K. Seppänen <jk...@ik...> wrote:

> sanders <sa...@kn...> writes:
>
> > If keywords fill=False and log=True,
> >
> > then after saving, the png looks fine but the histogram in the pdf is
> > mixed up.
>
> Confirmed, thanks for the report. I filed this at
> https://github.com/matplotlib/matplotlib/issues/804
>
>
I ran into something like this with filled plots originally saved as eps
files and then converted into a pdf.  Didn't need log=True, though.  The
eps files were fine, but depending on how I converted (pstopdf or ps2pdf, I
forget which) it would get messed up.  I will double-check my comments I
made in a Makefile about this to see if it might be related.

Ben Root

On Tue, Mar 27, 2012 at 5:39 PM, klo uo <kl...@gm...> wrote:

> Ah mayavi... I find it complicated for building, and in Ubuntu repository
> (or launchpad) there is some old version
> I'll try later today to build it
>
> Thanks for your suggestion,
> Cheers
>
>
>
> On Tue, Mar 27, 2012 at 11:25 PM, Frédéric Vogt <fv...@ms...>wrote:
>
>> Is the savefig function what you're looking for ?
>>
>> http://github.enthought.com/mayavi/mayavi/auto/mlab_figure.html
>>
>> Note that as far as I know, the .mtl file associated with a given .obj
>> file is not handled properly - all your colors will be white. Personally, I
>> export in .vrml which suits my needs better, as it does handle colors in a
>> satisfactory way.
>>
>>
>
With the way matplotlib is designed and structured, it may be a while
before that would be possible.  It wouldn't be impossible, mind you, but I
just haven't thought about that.  Could you file a wishlist item on
Github?  Maybe I could steal some code from Mayavi for this idea...

In the meantime, Mayavi would be your best bet, though.

Thanks,
Ben Root