matplotlib-users Mailing List for matplotlib (Page 11)

> I am not sure what you are asking -- could you elaborate?

Sorry.  I use matplotlib to create PNGs graphics for display on a
web-page.  I want to make the plots zoom-able.  I'll use javascript to
capture the pixel positions of the user's selected zoom region on the
PNG plot.  I'll then translate this into the coordinate system on the
plot and redraw it with a new x, y range.  I'm having trouble
translating the initial axises into the pixel positions (essentially
where they are on the PNG image).

For example, I use the following code to map point positions from the
coordinate system to pixel positions.  I use an img map <MAP> to
provide interaction with the actual lines.

[CODE]

lineObj = plt.plot(Xs,Ys,marker='o')[0]
path, affine = lineObj._transformed_path.get_transformed_points_and_affine()
path = affine.transform_path(path)
for real,pixel in zip(lineObj.get_xydata(),path.vertices):
 ## write <AREA> tag  for each point

[/CODE]

I'd like to get information similar to this for the axis of the plot.

Thanks.

Josh Lawrence wrote:
> Greetings all,
>
> In using the function griddata in mlab.py, I think I have found a bug.  
> The following line in mlab.py errors for me.
> I supply it an xi and yi that have shape (N,1). I have surface data,  
> but I only care about the variation in one direction. In mlab, when it  
> gets to this line (2956 in svn revision 7040):
>
>           if min(xo[1:]-xo[0:-1]) < 0 or min(yo[1:]-yo[0:-1]) < 0:
>               raise ValueError, 'output grid defined by xi,yi must be  
> monotone increasing'
>
> the result is an error. That is, I get the following:
>
>           ValueError: min() arg is an empty sequence
>
> A couple of things. First, if I make my variation in x to be 2 points  
> (x = 0 for the case I'm interested in--so I just have both values of x  
> be zero), I do not get this error and I believe the result works. So,  
> it seems that there should be some handling of the case that there are  
> only 1 point in either x or y direction.
>
> Second, is it better to use the builtin python function min, or should  
> numpy.min be used instead?
>
> Cheers,
>
> Josh Lawrence
> Ph.D. Student
> Clemson University

Josh:  griddata currently only works for 2-D output grids.  It may be 
possible to modify it to work with 1-D data, but that was not the 
original intent of the function.

-Jeff

-- 
Jeffrey S. Whitaker         Phone  : (303)497-6313
Meteorologist               FAX    : (303)497-6449
NOAA/OAR/PSD  R/PSD1        Email  : Jef...@no...
325 Broadway                Office : Skaggs Research Cntr 1D-113
Boulder, CO, USA 80303-3328 Web    : http://tinyurl.com/5telg

On Thu, Apr 16, 2009 at 1:10 PM, Mark Larsen <lar...@gm...> wrote:

> MPL Users,
>
> How can I get the matched max and min pixel/coordinate positions of my
> x and y axises?



I am not sure what you are asking -- could you elaborate?

JDH

Greetings all,

In using the function griddata in mlab.py, I think I have found a bug.  
The following line in mlab.py errors for me.
I supply it an xi and yi that have shape (N,1). I have surface data,  
but I only care about the variation in one direction. In mlab, when it  
gets to this line (2956 in svn revision 7040):

          if min(xo[1:]-xo[0:-1]) < 0 or min(yo[1:]-yo[0:-1]) < 0:
              raise ValueError, 'output grid defined by xi,yi must be  
monotone increasing'

the result is an error. That is, I get the following:

          ValueError: min() arg is an empty sequence

A couple of things. First, if I make my variation in x to be 2 points  
(x = 0 for the case I'm interested in--so I just have both values of x  
be zero), I do not get this error and I believe the result works. So,  
it seems that there should be some handling of the case that there are  
only 1 point in either x or y direction.

Second, is it better to use the builtin python function min, or should  
numpy.min be used instead?

Cheers,

Josh Lawrence
Ph.D. Student
Clemson University

MPL Users,

How can I get the matched max and min pixel/coordinate positions of my
x and y axises?

Thanks,

LarsenMTL

Dear matplotlib users/developers,

I recently noticed that the errorbar funciton in the pyplot module  
changed it's default setting to fmt="-".
This is a bit strange since this function is primarily used to display  
individual data points with measurement errors, drawing a line  
connecting the datapoints is not a common procedure and normally only  
done with a fit to the data.
I think it makes much more sense if the default behaviour was fmt="+".

Kind regards,

Pim Schellart

Images, one is directly png and other converted from eps.

Andres

Andres Luhamaa wrote:
> Hello!
> Following example will give strange result in postscript output. It does 
> not happen, if I use imshow or contourf instead of pcolormesh and it 
> seems to happen only with log scale.
>
> Best regards,
> Andres
>
> import numpy as np
> import matplotlib
> matplotlib.use( 'PS' )
> from mpl_toolkits.basemap import Basemap
> import os
> import pylab as plt
> import matplotlib.colors as colors
> m=Basemap(width=1500000,height=1000000,resolution='c',projection='stere',lon_0=\
> 17,lat_0=58)
> # geogr coords for 
> points                                                      
> rx=np.arange(18,22,0.1)
> ry=np.arange(56,60,0.1)
> RX,RY=np.meshgrid(rx,ry)
> x,y=m(RX,RY)
> # create data for 
> points                                                       
> data=np.zeros([rx.size,ry.size])
> data[:10,:10]=1e3
> m.pcolormesh(x,y,data
>              ,norm=colors.LogNorm(vmin=50,vmax=17000.0)
>              )
> m.drawcoastlines()
> m.drawcountries()
> m.drawmapboundary()
> plt.colorbar()
> plt.savefig('mapproblem.eps')
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> High Quality Requirements in a Collaborative Environment.
> Download a free trial of Rational Requirements Composer Now!
> http://p.sf.net/sfu/www-ibm-com
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>   

Hi, I want to put arrows onto the contours of a contour plot, the direction
will depend on whether the contour is positive/negative.  Before I try to do
it myself, has anybody written something to do that?

Thanks, Evan

Hi List,

AS> now you're not loading the
AS> graphical library that was grabbing focus, but merely a low(er)
AS> level drawing library.

This raises a question I've had for some time. I wish the figure method 
had x and y kwargs, so that I could specify where the figure window pops 
up. It always seems to pop up over my console, so that I have to move it 
to keep typing. It really breaks work flow. I also wish for something 
like a nofocus kwarg, that will tell the figure window to open in the 
background, so as not to grab focus. Again, this breaks work flow. This 
would be more like how Matlab figures behave. Although they popup to the 
foreground, you can continue typing and focus is correctly applied to 
the console.

-- 
Christopher Brown, Ph.D.
Department of Speech and Hearing Science
Arizona State University

Hello!
Following example will give strange result in postscript output. It does 
not happen, if I use imshow or contourf instead of pcolormesh and it 
seems to happen only with log scale.

Best regards,
Andres

import numpy as np
import matplotlib
matplotlib.use( 'PS' )
from mpl_toolkits.basemap import Basemap
import os
import pylab as plt
import matplotlib.colors as colors
m=Basemap(width=1500000,height=1000000,resolution='c',projection='stere',lon_0=\
17,lat_0=58)
# geogr coords for 
points                                                      
rx=np.arange(18,22,0.1)
ry=np.arange(56,60,0.1)
RX,RY=np.meshgrid(rx,ry)
x,y=m(RX,RY)
# create data for 
points                                                       
data=np.zeros([rx.size,ry.size])
data[:10,:10]=1e3
m.pcolormesh(x,y,data
             ,norm=colors.LogNorm(vmin=50,vmax=17000.0)
             )
m.drawcoastlines()
m.drawcountries()
m.drawmapboundary()
plt.colorbar()
plt.savefig('mapproblem.eps')

I am not able to reproduce this leak here with 0.98.6svn from today on 
RHEL4.  What platform are you on?

(See attached massif profile -- the memory usage is flat...)

Mike

Jesper Larsen wrote:
> Hi matplotlib developers and users,
>
> I have had some problems with a memory leak in a long running
> matplotlib based web application that I have developed
> (www.worldwildweather.com). I believe the problem is due to a memory
> leak in the Agg backend but I am not sure. Below is a script which for
> me results in a consistently increasing amount of  memory usage. I am
> using mpl version 0.98.6svn. The problem does not occur when the
> savefig command is commented out. And it does not occur when "cs =
> ax.contourf(z)" and "ax.cla()" are moved outside the loop (before and
> after respectively).
>
> Best regards,
> Jesper
>
> import os, gc
> import numpy as npy
> import matplotlib as mpl
> from matplotlib.figure import Figure
> from matplotlib.backends.backend_agg import FigureCanvasAgg as FigureCanvas
>
> def report_memory():
>     """Report memory."""
>     gc.collect()
>     pid = os.getpid()
>     a2 = os.popen('ps -p %d -o rss,vsz,%%mem' % pid).readlines()
>     return int(a2[1].split()[1])
>
> fig = Figure(dpi=100)
> ax = fig.add_axes([0.1, 0.1, 0.8, 0.8])
> FigureCanvas(fig)
>
> n = 1000
> z = npy.zeros((n,n))
> for i in range(2000):
>     cs = ax.contourf(z)
>     fig.savefig('test.png')
>     ax.cla()
>     print report_memory(), i
>
> I have not pasted in all of the output but just the first and last 25 lines:
> 53356 0
> 53360 1
> 53360 2
> 53360 3
> 53360 4
> 53360 5
> 53360 6
> 53360 7
> 53360 8
> 53360 9
> 53360 10
> 53360 11
> 53360 12
> 53360 13
> 53360 14
> 53360 15
> 53360 16
> 53360 17
> 53356 18
> 53360 19
> 53360 20
> 53360 21
> 53360 22
> 53360 23
> 53356 24
> ...
> 57552 1975
> 57552 1976
> 57552 1977
> 57552 1978
> 57552 1979
> 57552 1980
> 57552 1981
> 57552 1982
> 57552 1983
> 57552 1984
> 57552 1985
> 57552 1986
> 57552 1987
> 57552 1988
> 57552 1989
> 57552 1990
> 57552 1991
> 57552 1992
> 57552 1993
> 57552 1994
> 57552 1995
> 57552 1996
> 57552 1997
> 57552 1998
> 57552 1999
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> High Quality Requirements in a Collaborative Environment.
> Download a free trial of Rational Requirements Composer Now!
> http://p.sf.net/sfu/www-ibm-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

Jeff Whitaker wrote:
> Zane Selvans wrote:
>> Is there a method built into Numpy/SciPy or friends that will generate
>> a set of N points evenly (regularly - not randomly) sampling the
>> entire surface of a sphere?  I imagine people doing GCMs and other
>> geoscience in spherical coordinates have to do this pretty frequently,
>> so I'm sure someone's written it in Python somewhere.  My searches
>> aren't turning anything up immediately though.
>>
>> Thanks,
>> Zane
>>
>>   
> Zane:  This one has puzzled mathematicians for centuries - there is no 
> evenly spaced set of points on a sphere.  The golden section spiral (or 
> fibonacci) points are easy to code  (see code below), but don't form the 
> vertices of polygons.  For modelling, where you need the points to be 
> vertices of polygons, hexagonal icosahedral grid are typically used (see 
> http://kiwi.atmos.colostate.edu/BUGS/geodesic/), but they are tricky to 
> compute (I don't have any python code handy).

I'm attaching some code I wrote to subdivide and icosahedron for tiling
a sphere with hexagons (and the occasional pentagon).

-Andrew

Hi matplotlib developers and users,

I have had some problems with a memory leak in a long running
matplotlib based web application that I have developed
(www.worldwildweather.com). I believe the problem is due to a memory
leak in the Agg backend but I am not sure. Below is a script which for
me results in a consistently increasing amount of  memory usage. I am
using mpl version 0.98.6svn. The problem does not occur when the
savefig command is commented out. And it does not occur when "cs =
ax.contourf(z)" and "ax.cla()" are moved outside the loop (before and
after respectively).

Best regards,
Jesper

import os, gc
import numpy as npy
import matplotlib as mpl
from matplotlib.figure import Figure
from matplotlib.backends.backend_agg import FigureCanvasAgg as FigureCanvas

def report_memory():
    """Report memory."""
    gc.collect()
    pid = os.getpid()
    a2 = os.popen('ps -p %d -o rss,vsz,%%mem' % pid).readlines()
    return int(a2[1].split()[1])

fig = Figure(dpi=100)
ax = fig.add_axes([0.1, 0.1, 0.8, 0.8])
FigureCanvas(fig)

n = 1000
z = npy.zeros((n,n))
for i in range(2000):
    cs = ax.contourf(z)
    fig.savefig('test.png')
    ax.cla()
    print report_memory(), i

I have not pasted in all of the output but just the first and last 25 lines:
53356 0
53360 1
53360 2
53360 3
53360 4
53360 5
53360 6
53360 7
53360 8
53360 9
53360 10
53360 11
53360 12
53360 13
53360 14
53360 15
53360 16
53360 17
53356 18
53360 19
53360 20
53360 21
53360 22
53360 23
53356 24
...
57552 1975
57552 1976
57552 1977
57552 1978
57552 1979
57552 1980
57552 1981
57552 1982
57552 1983
57552 1984
57552 1985
57552 1986
57552 1987
57552 1988
57552 1989
57552 1990
57552 1991
57552 1992
57552 1993
57552 1994
57552 1995
57552 1996
57552 1997
57552 1998
57552 1999

On Wed, Apr 15, 2009 at 8:03 AM, Vito De Tullio <zak...@li...>wrote:

> Gregor Thalhammer wrote:
>
> > I didn't understand what you mean by 'plot nodes'
>
> the markers, sorry
>
> > , but there exist
> > figure_enter_event, figure_leave_event, axes_enter_event and
> > axes_leave_event, see
> >
>
> http://matplotlib.sourceforge.net/examples/event_handling/figure_axes_enter_leave.html
>
> first of all: why these *_enter_event and *_leave_event aren't in
>
> http://matplotlib.sourceforge.net/api/backend_bases_api.html#matplotlib.backend_bases.FigureCanvasBase.mpl_connect


This is just a doc oversight -- I'll fix it.

I think your solution is  the right way to do it with the exiting mpl
infrastructure, though one could use the "blit" api to make it a little more
efficient to redraw just the needed artists rather than the whole figure
each time (doing this would further complicate the code).  Since you call
your approach "ugly, and conceptually wrong", perhaps you can sketch what
you think the right interface should be for something like this, and we will
see if there is anything we can do within the existing mpl infrastructure to
make it easier.

FYI, check out the canvas.onHilite implementation, which shows how to use a
hover to highlight individual artists (title, axes, xlabel, ylabel,
ticklabel etc).  This problem is a little
easier, since what you are doing is triggering an action on artist B when
artist A is hovered over.

    import numpy as np
    import matplotlib.pyplot as plt


    fig = plt.figure()
    ax = fig.add_subplot(111)
    ax.plot(np.random.rand(10))
    ax.set_title('hover over me')
    ax.set_xlabel('or me')
    ax.set_ylabel('or me')
    ax.grid()
    fig.canvas.mpl_connect('motion_notify_event', fig.canvas.onHilite)

    plt.show()

JDH


JDH

Gregor Thalhammer wrote:

> I didn't understand what you mean by 'plot nodes'

the markers, sorry

> , but there exist 
> figure_enter_event, figure_leave_event, axes_enter_event and
> axes_leave_event, see
>
http://matplotlib.sourceforge.net/examples/event_handling/figure_axes_enter_leave.html

first of all: why these *_enter_event and *_leave_event aren't in
http://matplotlib.sourceforge.net/api/backend_bases_api.html#matplotlib.backend_bases.FigureCanvasBase.mpl_connect

second: they are cool, but I need to be more specific: I want to show some
annotations only if I move my mouse pointer over a marker, and to make them
disappear when I move my mouse pointer away (something like
http://www.scipy.org/Cookbook/Matplotlib/Interactive_Plotting, but on mouse
hovering)

-- 
By ZeD

Vito De Tullio schrieb:
> Hi all.
> I'm using matplotlib to draw some graph in a pyqt application.
> Is there a way to show infos only when the mouse hover the plot nodes?
> At the moment I'm stuck using this (ugly, and conceptually wrong) piece of
> code, written by me...
>
> basically I add hidden annotations where I need, then (ab)use the 
> fig.canvas.mpl_connect function with the 'motion_notify_event' option.
>
> Is there a "motion_mouse_enter_plot" / "motion_mouse_exit_plot" or something
> similar? I see there is the "pick_event" interface, but I don't want to
> show infos "on click", but "on hovering"
>   
I didn't understand what you mean by 'plot nodes', but there exist 
figure_enter_event, figure_leave_event, axes_enter_event and 
axes_leave_event, see
http://matplotlib.sourceforge.net/examples/event_handling/figure_axes_enter_leave.html

Gregor

Hi all.
I'm using matplotlib to draw some graph in a pyqt application.
Is there a way to show infos only when the mouse hover the plot nodes?
At the moment I'm stuck using this (ugly, and conceptually wrong) piece of
code, written by me...

basically I add hidden annotations where I need, then (ab)use the 
fig.canvas.mpl_connect function with the 'motion_notify_event' option.

Is there a "motion_mouse_enter_plot" / "motion_mouse_exit_plot" or something
similar? I see there is the "pick_event" interface, but I don't want to
show infos "on click", but "on hovering"

#!/usr/bin/env python

class show_annotation(object):
    def __init__(self, annotations):
        '''"fake" class, mostly a "function with memory"'''
        self.annotations = annotations

    def __call__(self, mouse_event):
        ax = mouse_event.inaxes
        if not ax:
            return
        line = ax.get_lines()[0]
        contained, infos = line.contains(mouse_event)
        if not contained:                                # eventually exited
            for annotation in self.annotations.values():
                annotation.set_visible(False)
        else:
            xdata, ydata = line.get_data()
            ind = infos['ind'][0]
            annotation = self.annotations[xdata[ind], ydata[ind]]
            if not annotation.get_visible():             # is entered
                annotation.set_visible(True)
        ax.figure.canvas.draw()

if __name__ == '__main__':
    from random import randint
    x = range(10)
    y = [ randint(0, i) for i in x ]
    labels = [ 'label_%d' % i for i in x ]

    from matplotlib import pyplot
    fig = pyplot.figure()
    ax = fig.add_subplot(1, 1, 1)
    ax.plot(x, y)

    # waiting for a "dict comprehension" sintax
    annotations = {}
    for label, xe, ye in zip(labels, x, y):
        annotations[xe,ye] = ax.annotate(label, (xe,ye), visible=False)

    fig.canvas.mpl_connect('motion_notify_event',
            show_annotation(annotations))
    pyplot.show()


-- 
By ZeD

Zane Selvans wrote:
> Is there a method built into Numpy/SciPy or friends that will generate
> a set of N points evenly (regularly - not randomly) sampling the
> entire surface of a sphere?  I imagine people doing GCMs and other
> geoscience in spherical coordinates have to do this pretty frequently,
> so I'm sure someone's written it in Python somewhere.  My searches
> aren't turning anything up immediately though.
>
> Thanks,
> Zane
>
>   
Zane:  This one has puzzled mathematicians for centuries - there is no 
evenly spaced set of points on a sphere.  The golden section spiral (or 
fibonacci) points are easy to code  (see code below), but don't form the 
vertices of polygons.  For modelling, where you need the points to be 
vertices of polygons, hexagonal icosahedral grid are typically used (see 
http://kiwi.atmos.colostate.edu/BUGS/geodesic/), but they are tricky to 
compute (I don't have any python code handy).

Regards,

-Jeff

import numpy as np
from mpl_toolkits.basemap import Basemap
import matplotlib.pyplot as plt
import sys

# from http://www.xsi-blog.com/archives/115

def fibonacci(N):
    inc = np.pi * (3 - np.sqrt(5))
    off = 2. / N
    r2d = 180./np.pi
    k = np.arange(0,N)
    y = k*off - 1. + 0.5*off
    r = np.sqrt(1 - y*y)
    phi = k * inc
    x = np.cos(phi)*r
    z = np.sin(phi)*r
    theta = np.arctan2(np.sqrt(x**2+y**2),z)
    phi = np.arctan2(y,x)
    lats = 90.-r2d*theta
    lons = r2d*phi
    return lats, lons

npts = int(sys.argv[1])

lats, lons = fibonacci(npts)

map = Basemap(projection ='ortho',lat_0=0,lon_0=-90)
map.drawcoastlines()
map.fillcontinents(color='coral')
map.drawmapboundary(fill_color='aqua')
x,y = map(lons, lats)
map.scatter(x,y,10,marker='o',zorder=10)
plt.show()

Thanks Jeff, seems better indeed with svn version, at least resolution 
'h' works (cant test the f resolution on a laptop). But when I zoom the 
image 4x, then gv starts giving errors again. At the same time, evince 
can show at the same zoom level.

Andres


Jeff Whitaker wrote:
> Andres Luhamaa wrote:
>> Hello!
>> I have a problem writing postscript output from a high resolution 
>> map. It does not give any error, but output file gives a lot of 
>> errors while I open it with gv. The following example works for 
>> basemap resolution 'i', but not with resolution 'f'. Any ideas what I 
>> should do in a different way or are there known limitations in PS or 
>> a bug?
>>
>> Andres
>>   
>
> Andres:  Works for me (using SVN trunk), so the bug in the ps backend 
> must have been fixed since the last release.
>
> -Jeff
>> #!/usr/bin/env 
>> python                                                          
>> import numpy as np
>> import matplotlib
>> matplotlib.use( 'PS' )
>> from mpl_toolkits.basemap import Basemap
>> import os
>> import matplotlib.pyplot
>> import pylab as plt
>> import matplotlib.colors as colors
>> fig=plt.figure()
>> m=Basemap(llcrnrlon=0.83513989,llcrnrlat=53.49684419,urcrnrlon=55.00709304,urcr\ 
>>
>> nrlat=64.88386147,
>>         projection='lcc',lat_0=60,lon_0=0.,
>>           resolution ='i'
>> #          
>> ,area_thresh=100.                                                   
>>           )
>> m.drawcoastlines()
>> m.drawcountries()
>> m.drawmapboundary()
>> m.drawparallels([40,50,60,70],labels=[1,0,0,1])
>> m.drawmeridians([-160,-140,-120,-100,-80,-60,-40,-20,0,20,40,60,80,100,120,140,\ 
>>
>> 160,180],labels=[0,1,0,1])
>> plt.savefig('surfgeopot.eps')
>>
>>
>> ------------------------------------------------------------------------------ 
>>
>> This SF.net email is sponsored by:
>> High Quality Requirements in a Collaborative Environment.
>> Download a free trial of Rational Requirements Composer Now!
>> http://p.sf.net/sfu/www-ibm-com
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>   
>
>

Hello,I am relatively new to matplotlib but I get the basics.  What I need
to do though is a slightly customized graph area.  There are two plots on
the same graph - a parabola and a straight line (with a negative slope)
north east of the parabola.  This graph has grid lines.  Simple enough, but
now I want to remove the graph area to the right of the line.  Think of a
dog-eared page in a book.  Basically that top right triangle is gone.

Can I do this in MPL or do I need to do some post processing in PIL?  Much
thanks for any help.

-Andrew

Is there a method built into Numpy/SciPy or friends that will generate
a set of N points evenly (regularly - not randomly) sampling the
entire surface of a sphere?  I imagine people doing GCMs and other
geoscience in spherical coordinates have to do this pretty frequently,
so I'm sure someone's written it in Python somewhere.  My searches
aren't turning anything up immediately though.

Thanks,
Zane

-- 
Zane A. Selvans
Amateur Earthling
http://zaneselvans.org
+1 303 815 6866

Tyler B wrote:
> I think that did it --- thanks Andrew!!
> 
> Just out of curiosity, what does 'Agg' refer to?

It's the low-level drawing library, anti-grain. You're bypassing the
loading of whatever other "backend" -- a low-level drawing system --
that matplotlib was loading before.

 Also, I'm guessing
> that some efficiency is lost once we make Python import the entire
> library (as opposed to just pyplot as before) -- is there a more
> efficient way, or is that just the price to pay?

Actually, I'd guess it's the opposite -- now you're not loading the
graphical library that was grabbing focus, but merely a low(er) level
drawing library.

-Andrew

I think that did it --- thanks Andrew!!

Just out of curiosity, what does 'Agg' refer to? Also, I'm guessing
that some efficiency is lost once we make Python import the entire
library (as opposed to just pyplot as before) -- is there a more
efficient way, or is that just the price to pay?

Either way, thanks a ton!
Tyler

On 4/15/09, Andrew Straw <str...@as...> wrote:
> Tyler B wrote:
>> Hi there,
>>
>> I am trying to run a python script on my computer at a regular
>> interval [in the background] to output an image file using
>> matplotlib.  My problem is that every time the script runs, it
>> momentarily steals focus from whatever I'm working on -- which over
>> time gets to be very annoying.
>>
>> I think that I've narrowed it down to a single line in the python script:
>> fig = plt.figure(facecolor ='w', frameon=False)
>>
>> [note that I've imported matplotlib.pyplot as plt earlier]
>>
>> I made a short video to show the problem:
>> http://screencast.com/t/MRObeKrx
>>
>> Can anyone think of a good way to prevent matplotlib from stealing
>> focus??  This is the last problem I have to solve.. though I'm out of
>> ideas.
> You don't show the beginning of your script, but try:
>
> import matplotlib
> matplotlib.use('Agg')
>
> before any other MPL imports.
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> High Quality Requirements in a Collaborative Environment.
> Download a free trial of Rational Requirements Composer Now!
> http://p.sf.net/sfu/www-ibm-com
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

Tyler B wrote:
> Hi there,
>
> I am trying to run a python script on my computer at a regular
> interval [in the background] to output an image file using
> matplotlib.  My problem is that every time the script runs, it
> momentarily steals focus from whatever I'm working on -- which over
> time gets to be very annoying.
>
> I think that I've narrowed it down to a single line in the python script:
> fig = plt.figure(facecolor ='w', frameon=False)
>
> [note that I've imported matplotlib.pyplot as plt earlier]
>
> I made a short video to show the problem:
> http://screencast.com/t/MRObeKrx
>
> Can anyone think of a good way to prevent matplotlib from stealing
> focus??  This is the last problem I have to solve.. though I'm out of
> ideas.
You don't show the beginning of your script, but try:

import matplotlib
matplotlib.use('Agg')

before any other MPL imports.

Thanks, Ryan, John, and Pierre... I will try to change it at the
point John suggests, maybe inspired by scikits.timeseries
(or just using it).

btw, I think adding the ability to set the AutoDateFormat
formatting choices per scale would be a good small addition
to mpl.  (Maybe  that is what the commented lines in the
class are about, but I don't quite understand them).  Methods
like  .set_dayscale_format, .set_weekscale_format, etc.

Thanks,
Che