matplotlib-users Mailing List for matplotlib

On Fri, Feb 6, 2009 at 5:13 AM, Rezwan <rez...@gm...> wrote:
>
> Hi
> I am developing a plotting tool using matplotlib and wxpython as backend. I
> have separate thread for iperf and it creates a text file every time I want
> to create this thread and this thread finish it's task once it reach the
> duration(defined by the user). Plot and some text boxes are updated reading
> this text file every second once I press plot button. My problem is
> matplotlib keeps the old data and impose new data on the same plot. I am
> using cla function to clear the axes but how can I clear set_data for every
> new plot. What I want is to initialize set_data. Anyone here to tell me how
> can I do that? Hope my post makes sense. If not I will post with some
> snippets of my tool.
>

You will need to post some example code before we can really help, but
the best way to do this is not via cla, but by storing the line
instance and calling set_data on it

  line, = ax.plot(x, y)

and then later

  line.set_data(newx, newy)
  line.axes.figure.canvas.draw()

If you are getting overplotting, it sounds like you are making
additional calls to plot that you are overlooking.

JDH

On Fri, Feb 6, 2009 at 8:34 AM, Gary Ruben <gr...@bi...> wrote:
> Hi all,
>
> I've attached a candidate imsave() to complement imread() in the image.py
> module. Would my use of pyplot instead of the oo interface preclude its
> inclusion in image.py? Also, I noticed some problems when I ran the tests
> with the Wx backends with mpl 0.98.5.2 in Win32. Both of the Wx backends
> produce incorrect, but different, results.

Yes, if it imports pyplot, we can't put it in the image module.  But
you can easily rewrite your function to use the Agg backend as in

  http://matplotlib.sourceforge.net/examples/api/agg_oo.html

the agg backend will detect other extensions, like png, pdf, svg or ps
and do the right thing.  So if you'd like to resubmit this as an svn
patch that doesn't use pyplot, I think it would be a useful addition.

  http://matplotlib.sourceforge.net/faq/howto_faq.html#contributing-howto

JDH

Eli Bressert wrote:
> Hi Everyone,
> 
> Looks like I may have run into a bug for the contourf function. I was  
> able to reproduce the problem on two OS X systems. One was based on  
> 10.5 and the other was on 10.4. The problem appears when you use  
> contourf with alpha < 1. With the transparency there appears to be  
> streaks of lines pointing downward from the contour lines. Is this a  
> bug that has been spotted before? Additional information is provided  
> below with a python script to reproduce the problem.
> 
> Note, this bug was reproduced with a range of different parameters and  
> input values. The script is the easiest way to reproduce the problem.

It is partly inherent in the underlying contouring algorithm, and I 
think partly reflecting a common characteristic of renderers.  On my 
ubuntu box, the problem shows up in agg, pdf or ps shown with evince, 
but *not* in svg rendered by eog.

The part inherent in the contouring algorithm is that all patches are 
simply connected--the algorithm does not make annular patches, for 
example--so there is a vertical cut.  With alpha < 1, that cut, and for 
that matter the boundary between one patch and the next, seem to be 
effectively rendered twice.  The contour code is already specifying that 
the patch should be rendered without a boundary, so I don't know what 
else can be done.

Eric
> 
> Cheers,
> 
> Eli
> 
> 
> 
> Mac OSX Darwin Kernel Version 9.5.0  (Leopard 10.5.5)
> 
> Matplotlib version: 0.98.3
> 
> Matplotlib was installed via EPD, version Py2.5 4.1.30101 i386
> 
> Code to show bug:
> Most of the python code was borrowed from the Matplotlib examples
> http://matplotlib.sourceforge.net/examples/pylab_examples/contour_demo.html?highlight=contours
> 
> ###### Begin Python Code #######
> import matplotlib
> import numpy as np
> import matplotlib.cm as cm
> import matplotlib.mlab as mlab
> import matplotlib.pyplot as plt
> 
> delta = 0.025
> x = np.arange(-3.0, 3.0, delta)
> y = np.arange(-2.0, 2.0, delta)
> X, Y = np.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)
> Z = 10.0 * (Z2 - Z1)
> 
> plt.figure()
> CS = plt.contourf(X, Y, Z,alpha = 0.7)
> ###### End Python Code #######
> 
> 
> 
> 
> 
> 
> 
> 
> 
> ------------------------------------------------------------------------------
> Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
> software. With Adobe AIR, Ajax developers can use existing skills and code to
> build responsive, highly engaging applications that combine the power of local
> resources and data with the reach of the web. Download the Adobe AIR SDK and
> Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Hi Everyone,

Looks like I may have run into a bug for the contourf function. I was  
able to reproduce the problem on two OS X systems. One was based on  
10.5 and the other was on 10.4. The problem appears when you use  
contourf with alpha < 1. With the transparency there appears to be  
streaks of lines pointing downward from the contour lines. Is this a  
bug that has been spotted before? Additional information is provided  
below with a python script to reproduce the problem.

Note, this bug was reproduced with a range of different parameters and  
input values. The script is the easiest way to reproduce the problem.

Cheers,

Eli



Mac OSX Darwin Kernel Version 9.5.0  (Leopard 10.5.5)

Matplotlib version: 0.98.3

Matplotlib was installed via EPD, version Py2.5 4.1.30101 i386

Code to show bug:
Most of the python code was borrowed from the Matplotlib examples
http://matplotlib.sourceforge.net/examples/pylab_examples/contour_demo.html?highlight=contours

###### Begin Python Code #######
import matplotlib
import numpy as np
import matplotlib.cm as cm
import matplotlib.mlab as mlab
import matplotlib.pyplot as plt

delta = 0.025
x = np.arange(-3.0, 3.0, delta)
y = np.arange(-2.0, 2.0, delta)
X, Y = np.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)
Z = 10.0 * (Z2 - Z1)

plt.figure()
CS = plt.contourf(X, Y, Z,alpha = 0.7)
###### End Python Code #######