matplotlib-users Mailing List for matplotlib

On Fri, 2006-10-13 at 13:59 -0400, Hubert Fitch wrote:
> Hi Everyone!
>  
> I have now been able to use Matplotlib on Window XP platfrom! Thanks
> for thiose who helped me to get Enthought Edition installed and
> working!
>  
> I have no idea how to do what I have wanted to do for several years. I
> will try to describe the plot I need.
> I need to be able to include this plot in a Wikki.
>  
> A point on a small circle with radius r0, is spiralling around
> (orthogonally) in a larger circle of radius r1, with these two radii
> related by r0/r1 = alpha. The smaller circle sweeps around the r1
> radius to create a path, with the point (on r1) locus describing a
> torroid coil in the surface of the ring torus. 
>  
> This describes the plot when the torus is at-rest, but when the ring
> torus is in linear motion, both r0 and r1 change dimensions together,
> (keeping the same r0/r1 = alpha ratio), to create an increasing or
> decreasing spiralling circular helix, that is a fuction of the mass
> energy created by this motion. The point on the r0 circle is "like a
> snake chasing its tail?"
>  
> When the torus is at-rest, the toroid path is closed, but in-motion,
> the torroid is open and becomes an increasing or decreasing spiral.
> This path will collapse back to a larger size if the energy is
> decreasing, and shrink to a smaller size if the energy is increasing.
>  
> If the at-rest motion plot can be created ( I know that I can't do
> this plot by myself), then an animation of the ring torus figure
> moving through space to create a spiral would help in visualizing this
> figure. If this animation can slow down or speed up this would be
> great..

Hmmm interesting ....

It's hard to follow your text description, but I think what you're
describing is at least a 3D visualisation, so matplotlib isn't ideal for
the task. Can you put this into equations? I'm not clear if your at-rest
plot should be a "trajectory" for a point moving round the torus

If you're willing to learn a new toolkit, this is a good task for VTK
(www.vtk.org). It's included in the enthought python distribution. The
problem here is that you almost certainly need to buy the user-guide
and/or the VTK text book to learn how to use it. VTK is very powerful,
but rather complex to learn.

Maybe this is a good task for VPython (http://www.vpython.org/ ). I've
never used this myself, but if looks well suited to this kind of dynamic
visualisation and is less intimidating than VTK (although it's built on
VTK I think).

Well below is a script to render a "coil" in 3D. It's too much work for
me to figure out the vector/matrix algebra to calculate the trajectory
you describe, but hopefully you get the idea: replace my 'trajectory'
function with whatever is appropriate for what you want. Maybe this will
get you started.

HTH
Bryan

-----------------------------------------------------------------------

"""
Draws a coil in 3D.

Mouse click to rotate the view.
Hit 'u' to start the animation (kill the program to stop it!)
"""

import vtk
import math
import time

offset = 0.0 #I'm going to use this to 'animate' the figure

def trajectory(t):
    """
    returns 3D positions, as a function of one parameter, t
    """
    global offset
    r1 = 1.0
    r0 = 0.2
    omega=5.0
    z = r1*math.sin(omega*t+offset)
    y = r1*math.cos(omega*t+offset)
    x = 0.5*t
    return (x,y,z)


s = vtk.vtkProgrammableSource()
def Execute():
    """The Execute method to be called by the source object"""
    #make a range of parameter values
    tVals = numpy.arange(0,2*math.pi,0.05)
    #convert these to positions
    positions = [trajectory(t) for t in tVals]
    
    #construct a VTK polydata object: a single poly-line 'cell'
    dataset = s.GetPolyDataOutput()
    points = vtk.vtkPoints()
    cells = vtk.vtkCellArray()
    cells.InsertNextCell(len(positions))
    for p in positions:
        id = points.InsertNextPoint(*p)
        cells.InsertCellPoint(id)
    dataset.SetPoints(points)
    dataset.SetLines(cells)
    return dataset
s.SetExecuteMethod(Execute)

#convert (poly)lines to a tube
t = vtk.vtkTubeFilter()
t.SetInput(s.GetOutput())
t.SetRadius(0.05)
t.SetNumberOfSides(10)

#make surface normals, so it renders smooth
n = vtk.vtkPolyDataNormals()
n.SetInput(t.GetOutput())

#boiler-plate VTK stuff....
m=vtk.vtkPolyDataMapper()
m.SetInput(n.GetOutput())

a=vtk.vtkActor()
a.SetMapper(m)

ren = vtk.vtkRenderer()
ren.AddActor(a)

renwin = vtk.vtkRenderWindow()
renwin.AddRenderer(ren)

iren = vtk.vtkRenderWindowInteractor()
iren.SetRenderWindow(renwin)

#add a simple animation routine, activated by the 'u' keypress
def Animate(obj, evt):
    global offset
    while 1: #need to kill the program to quit
        offset += 0.1
        s.Modified()
        renwin.Render()
        time.sleep(0.05)
iren.AddObserver("UserEvent", Animate)
iren.Initialize()
iren.Start()


>  
> Can anyone help me to create this plot?
>  
> Thanks for any help!
>  
> Hubert fitch
>  
>  
> -------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________ Matplotlib-users mailing list Mat...@li... https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Hello

After compiling and installing matplotlib-0.87.6, when I entered,

from pylab import *
it returned the following errors.

Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/home/morovia/bin/lib/python2.4/site-packages/pylab.py", line 1, in
?
    from matplotlib.pylab import *
  File"/home/morovia/bin/lib/python2.4/site-packages/matplotlib/pylab.py",
line220, in ?
    new_figure_manager, draw_if_interactive, show = pylab_setup()
  File
"/home/morovia/bin/lib/python2.4/site-packages/matplotlib/backends/__init__.py",
line 23, in pylab_setup
    globals(),locals(),[backend_name])
  File
"/home/morovia/bin/lib/python2.4/site-packages/matplotlib/backends/backend_tkagg.py",
line 8, in ?
    import tkagg                 # Paint image to Tk photo blitter extension
  File
"/home/morovia/bin/lib/python2.4/site-packages/matplotlib/backends/tkagg.py",
line 1, in ?
    import _tkagg
ImportError: matplotlib/backends/_tkagg.so: undefined symbol: XFreePixmap
********************************************************************************************************************************
looks like I need to include X11 libraries,

Kindly inform me where I need to modify.

Thanks
Best regards
Morovia.

Hi Everyone!

I have now been able to use Matplotlib on Window XP platfrom! Thanks for =
thiose who helped me to get Enthought Edition installed and working!

I have no idea how to do what I have wanted to do for several years. I =
will try to describe the plot I need.
I need to be able to include this plot in a Wikki.

A point on a small circle with radius r0, is spiralling around =
(orthogonally) in a larger circle of radius r1, with these two radii =
related by r0/r1 =3D alpha. The smaller circle sweeps around the r1 =
radius to create a path, with the point (on r1) locus describing a =
torroid coil in the surface of the ring torus.=20

This describes the plot when the torus is at-rest, but when the ring =
torus is in linear motion, both r0 and r1 change dimensions together, =
(keeping the same r0/r1 =3D alpha ratio), to create an increasing or =
decreasing spiralling circular helix, that is a fuction of the mass =
energy created by this motion. The point on the r0 circle is "like a =
snake chasing its tail?"

When the torus is at-rest, the toroid path is closed, but in-motion, the =
torroid is open and becomes an increasing or decreasing spiral. This =
path will collapse back to a larger size if the energy is decreasing, =
and shrink to a smaller size if the energy is increasing.

If the at-rest motion plot can be created ( I know that I can't do this =
plot by myself), then an animation of the ring torus figure moving =
through space to create a spiral would help in visualizing this figure. =
If this animation can slow down or speed up this would be great..

Can anyone help me to create this plot?

Thanks for any help!

Hubert fitch

Hanno,
I wrestled with this one for a while. The attached script is the result
of my effort. The two keys were to define the colormap, and then make
sure that when you call the colormap that the values span the full
range. For example, if you want a data point to be "yellow", and that
color (RGB: 1,1,0) is between 0.75 and 0.85 on your colormap, then give
the point a value of 0.81 (or whatever in the range). Then be sure to
also pass a zero (0.0) and a one (1.0) to the scatter. By spanning the
full scale range you will be sure to fix the color that you want from
the colormap to the data point that you define. It might be a hack, but
it works ok.

Cheers,
Jake

-----Original Message-----
From: Louis Pecora [mailto:pe...@an...]=20
Sent: Friday, October 13, 2006 7:47 AM
To: mat...@li...
Subject: Re: [Matplotlib-users] plot x y value(x,y)

Hanno Klemm wrote:
> Hi Louis,
>
> sorry for being unclear. What I have is a list of data points given in

> the format
>
> x y value
>
> so my array looks like
>
> x_1, y_1, v_1
> x_2, y_2, v_2,
>
> etc.
>
> Now I want to plot a point at (x_i, y_i) and assign to it a color=20
> according to v_i. The problem is, that the values at x and y are given

> and are not calculated.
>
> Christian's suggestion with a scatter plot works fine for me, but if=20
> there are other possibilities I'm always keen to learn something.
>
> Hope this clarifies matters.
>
> Best regards,
> Hanno
>
>  =20

Got it.  Yes, sounds like you want a "3D" scatter plot.  v (at some
"height) over the surface of x,y points.  There is a contour in
matplotlib, but I don't think it will do what you want.  I will go back
and look over matplotlib.

You might want to go to the matplotlib web site and download the
examples.  There are tons of them.  Maybe one is right for you.

--
Cheers,

Lou Pecora

Code 6362
Naval Research Lab
Washington, DC  20375
USA
Ph:  +202-767-6002
email:  pe...@an...

Hanno Klemm wrote:
> Louis,
>
> my problem has been solved with your hint to contour and contourf.
>
> Thanks again,
> Hanno
>
>
>
>   

Glad that helped although I wasn't sure.

Your post did make me realize that a 3D scatter plot with option to 
connect points is a good idea.  In my quick glance at the examples, I 
didn't see it, but I often have data that I want to plot as 3D points or 
line segments, i.e. as parametric plots, e.g. x,y,z positions of an 
object as a function of time.

I'll have to look over matplotlib to see if this is doable.  It would be 
nice.

-- 
Cheers,

Lou Pecora

Code 6362
Naval Research Lab
Washington, DC  20375
USA
Ph:  +202-767-6002
email:  pe...@an...

>>>>> "Ryan" == Ryan May <rm...@ou...> writes:


    Ryan> Looking at the FigureImage class in image.py, make_image
    Ryan> doesn't seem to have any clue about the magnification
    Ryan> parameter.  This call is in figure.py, and not the demo
    Ryan> code, so it would appear to be a bug in the library.  Any
    Ryan> idea what's wrong?

Should be fixed in svn.

JDH

Hi,

I've been playing with the matplotlib demos (very impressive btw.) and 
I've found that figimage_demo.py and pcolor_nonuniform.py give 
exceptions on my computer.  Here's the last part of the backtrace for 
pcolor_nonuniform.py:

Traceback (most recent call last):
   File 
"/usr/lib64/python2.4/site-packages/matplotlib/backends/backend_gtk.py", 
line 284, in expose_event
     self._render_figure(self._pixmap, w, h)
   File 
"/usr/lib64/python2.4/site-packages/matplotlib/backends/backend_gtkagg.py", 
line 73, in _render_figure
     FigureCanvasAgg.draw(self)
   File 
"/usr/lib64/python2.4/site-packages/matplotlib/backends/backend_agg.py", 
line 392, in draw
     self.figure.draw(renderer)
   File "/usr/lib64/python2.4/site-packages/matplotlib/figure.py", line 
545, in draw
     for a in self.axes: a.draw(renderer)
   File "/usr/lib64/python2.4/site-packages/matplotlib/axes.py", line 
1067, in draw
     a.draw(renderer)
   File "/usr/lib64/python2.4/site-packages/matplotlib/image.py", line 
189, in draw
     im = self.make_image(renderer.get_image_magnification())
TypeError: make_image() takes exactly 1 argument (2 given)


Looking at the FigureImage class in image.py, make_image doesn't seem to 
have any clue about the magnification parameter.  This call is in 
figure.py, and not the demo code, so it would appear to be a bug in the 
library.  Any idea what's wrong?

Thanks,

Ryan

-- 
Ryan May
Graduate Research Assistant
School of Meteorology
University of Oklahoma
rm...@ro...

Yes, actually that's even better for what I want. However, after
manipulating the Z values into a 2d array, I can then just as well use
imshow to view their values.

So I'll stick to the scatter plot approach for unevenly spaced data
and to the contour/imshow approach for evenly gridded data.

Thanks again,

Hanno



Christian Meesters <mee...@un...> said:

> I think Louis' idea of using contour plots for your problem is not
that bad. 
> It depends, of course, on the actual meaning of your data ...
> 
> Assume you understand German? Have a look here: 
> http://www.python-forum.de/topic-5294.html
> 
> Christian
> 
> On Friday 13 October 2006 15:19, Hanno Klemm wrote:
> > Hi Louis,
> >
> > sorry for being unclear. What I have is a list of data points given in
> > the format
> >
> > x y value
> >
> > so my array looks like
> >
> > x_1, y_1, v_1
> > x_2, y_2, v_2,
> >
> > etc.
> >
> > Now I want to plot a point at (x_i, y_i) and assign to it a color
> > according to v_i. The problem is, that the values at x and y are given
> > and are not calculated.
> >
> > Christian's suggestion with a scatter plot works fine for me, but if
> > there are other possibilities I'm always keen to learn something.
> >
> > Hope this clarifies matters.
> >
> > Best regards,
> > Hanno
> >
> > Louis Pecora <pe...@an...> said:
> > > Christian Meesters wrote:
> > > > Hi,
> > > >
> > > > No stupid question at all! I don't really understand your
> >
> > question, but it
> >
> > > > sounds to me like a scatter plot is what you want. In that case,
> >
> > just have a
> >
> > > > look at the scatter plot demos on the web page (-> screenshots) or
> >
> > in the
> >
> > > > example files.
> > > >
> > > > HTH
> > > > Christian
> > >
> > > Actually sounds more like a surface or contour plot.  You have x, y,
> >
> > and
> >
> > > z where z=function of (x,y).  I thought there was a contour call in
> > > matplotlib.  I'll look.
> > >
> > > --
> > > Cheers,
> > >
> > > Lou Pecora
> > >
> > > Code 6362
> > > Naval Research Lab
> > > Washington, DC  20375
> > > USA
> > > Ph:  +202-767-6002
> > > email:  pe...@an...
> >
> >
-------------------------------------------------------------------------
> >
> > > Using Tomcat but need to do more? Need to support web services,
> >
> > security?
> >
> > > Get stuff done quickly with pre-integrated technology to make your
> >
> > job easier
> >
> > > Download IBM WebSphere Application Server v.1.0.1 based on Apache
> >
> > Geronimo
> >
> > >
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> > > _______________________________________________
> > > Matplotlib-users mailing list
> > > Mat...@li...
> > > https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
>
-------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services,
security?
> Get stuff done quickly with pre-integrated technology to make your
job easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache
Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 



-- 
Hanno Klemm
kl...@ph...

Hi all,
after some suggestions, here is the latest version of windrose. Now, keywords 
args are passed by **kwargs, "bar" style have a "opening" arg for controlling 
the sectors angles, legends are positionned with "legendloc" and their size 
with "legendsize", and box are not displayed anymore.
Hope you find it usefull, any comments welcome.

Cordialement

-- 
Lionel Roubeyrie - lro...@li...
LIMAIR
http://www.limair.asso.fr

Hanno Klemm wrote:
> Hi Louis,
>
> sorry for being unclear. What I have is a list of data points given in
> the format
>
> x y value
>
> so my array looks like
>
> x_1, y_1, v_1
> x_2, y_2, v_2,
>
> etc.
>
> Now I want to plot a point at (x_i, y_i) and assign to it a color
> according to v_i. The problem is, that the values at x and y are given
> and are not calculated.
>
> Christian's suggestion with a scatter plot works fine for me, but if
> there are other possibilities I'm always keen to learn something.
>
> Hope this clarifies matters.
>
> Best regards,
> Hanno
>
>   

Got it.  Yes, sounds like you want a "3D" scatter plot.  v (at some 
"height) over the surface of x,y points.  There is a contour in 
matplotlib, but I don't think it will do what you want.  I will go back 
and look over matplotlib.

You might want to go to the matplotlib web site and download the 
examples.  There are tons of them.  Maybe one is right for you.

-- 
Cheers,

Lou Pecora

Code 6362
Naval Research Lab
Washington, DC  20375
USA
Ph:  +202-767-6002
email:  pe...@an...

I think Louis' idea of using contour plots for your problem is not that bad. 
It depends, of course, on the actual meaning of your data ...

Assume you understand German? Have a look here: 
http://www.python-forum.de/topic-5294.html

Christian

On Friday 13 October 2006 15:19, Hanno Klemm wrote:
> Hi Louis,
>
> sorry for being unclear. What I have is a list of data points given in
> the format
>
> x y value
>
> so my array looks like
>
> x_1, y_1, v_1
> x_2, y_2, v_2,
>
> etc.
>
> Now I want to plot a point at (x_i, y_i) and assign to it a color
> according to v_i. The problem is, that the values at x and y are given
> and are not calculated.
>
> Christian's suggestion with a scatter plot works fine for me, but if
> there are other possibilities I'm always keen to learn something.
>
> Hope this clarifies matters.
>
> Best regards,
> Hanno
>
> Louis Pecora <pe...@an...> said:
> > Christian Meesters wrote:
> > > Hi,
> > >
> > > No stupid question at all! I don't really understand your
>
> question, but it
>
> > > sounds to me like a scatter plot is what you want. In that case,
>
> just have a
>
> > > look at the scatter plot demos on the web page (-> screenshots) or
>
> in the
>
> > > example files.
> > >
> > > HTH
> > > Christian
> >
> > Actually sounds more like a surface or contour plot.  You have x, y,
>
> and
>
> > z where z=function of (x,y).  I thought there was a contour call in
> > matplotlib.  I'll look.
> >
> > --
> > Cheers,
> >
> > Lou Pecora
> >
> > Code 6362
> > Naval Research Lab
> > Washington, DC  20375
> > USA
> > Ph:  +202-767-6002
> > email:  pe...@an...
>
> -------------------------------------------------------------------------
>
> > Using Tomcat but need to do more? Need to support web services,
>
> security?
>
> > Get stuff done quickly with pre-integrated technology to make your
>
> job easier
>
> > Download IBM WebSphere Application Server v.1.0.1 based on Apache
>
> Geronimo
>
> > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> > _______________________________________________
> > Matplotlib-users mailing list
> > Mat...@li...
> > https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Hi Louis,

sorry for being unclear. What I have is a list of data points given in
the format

x y value

so my array looks like

x_1, y_1, v_1
x_2, y_2, v_2,

etc.

Now I want to plot a point at (x_i, y_i) and assign to it a color
according to v_i. The problem is, that the values at x and y are given
and are not calculated.

Christian's suggestion with a scatter plot works fine for me, but if
there are other possibilities I'm always keen to learn something.

Hope this clarifies matters.

Best regards,
Hanno


Louis Pecora <pe...@an...> said:

> Christian Meesters wrote:
> > Hi,
> >
> > No stupid question at all! I don't really understand your
question, but it 
> > sounds to me like a scatter plot is what you want. In that case,
just have a 
> > look at the scatter plot demos on the web page (-> screenshots) or
in the 
> > example files.
> >
> > HTH
> > Christian
> >
> >   
> 
> Actually sounds more like a surface or contour plot.  You have x, y,
and 
> z where z=function of (x,y).  I thought there was a contour call in 
> matplotlib.  I'll look.
> 
> -- 
> Cheers,
> 
> Lou Pecora
> 
> Code 6362
> Naval Research Lab
> Washington, DC  20375
> USA
> Ph:  +202-767-6002
> email:  pe...@an...
> 
> 
>
-------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services,
security?
> Get stuff done quickly with pre-integrated technology to make your
job easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache
Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 



-- 
Hanno Klemm
kl...@ph...

Christian Meesters wrote:
> Hi,
>
> No stupid question at all! I don't really understand your question, but it 
> sounds to me like a scatter plot is what you want. In that case, just have a 
> look at the scatter plot demos on the web page (-> screenshots) or in the 
> example files.
>
> HTH
> Christian
>
>   

Actually sounds more like a surface or contour plot.  You have x, y, and 
z where z=function of (x,y).  I thought there was a contour call in 
matplotlib.  I'll look.

-- 
Cheers,

Lou Pecora

Code 6362
Naval Research Lab
Washington, DC  20375
USA
Ph:  +202-767-6002
email:  pe...@an...

Hi,

No stupid question at all! I don't really understand your question, but it 
sounds to me like a scatter plot is what you want. In that case, just have a 
look at the scatter plot demos on the web page (-> screenshots) or in the 
example files.

HTH
Christian

On Friday 13 October 2006 14:25, Hanno Klemm wrote:
> Hi,
>
> this is probably a stupid question, but somehow I just can't figure
> out how to do it. I have data given in an array with entries:
>
> x y data(x,y)
>
> and thus of shape (#data points, 3)
>
> what is the easiest way with mpl to plot these data points, susch that
> they are located at x,y and colored according to data(xy)?
>
> Thanks for any pointers,
> Hanno

Hi,

this is probably a stupid question, but somehow I just can't figure
out how to do it. I have data given in an array with entries:

x y data(x,y)

and thus of shape (#data points, 3)

what is the easiest way with mpl to plot these data points, susch that
they are located at x,y and colored according to data(xy)?

Thanks for any pointers,
Hanno


-- 
Hanno Klemm
kl...@ph...

Edit your "~/.matplotlib/matplotlibrc" file to specify the default.
On OSX WXAgg or TkAgg is the best bet.

On 10/12/06, Louis Pecora <pe...@an...> wrote:
>
> In order get pylab to work on my Mac I have to do the following (found
> by some experimenting):
>
> import matplotlib
> matplotlib.use("WXAgg")
> import pylab
>
> Does this seem right?  I mean, it works, but it seems to be a little
> gymnastics to go through rather than just doing
>
> import pylab
>
> which gives
>
> ImportError: No module named gobject
>
> since it can't seem to find that backend in the "backend" directory when
> it tries to import backend_gtkagg.py.  I guess that doesn't work on the
> Mac.
>
>
> My setup:   (Tiger 10.4) Al Laptop, Python 2.4, latest pylab from SciPy
> package.
>
> --
> Cheers,
>
> Lou Pecora
>
> Code 6362
> Naval Research Lab
> Washington, DC  20375
> USA
> Ph:  +202-767-6002
> email:  pe...@an...
>
>
> -------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

Le vendredi 13 octobre 2006 08:26, Derek a =E9crit=A0:
> Lionel Roubeyrie <lroubeyrie@...> writes:
> > Hi all,
> > continuing on the windroses, here the third version of windrose.py, with
> > two others styles (line and bar2 (is it what you want Derek?)). I want =
to
> > know how it's possible to modify the legends to be on axes, and not on
> > the figure like in the ex4.png subplots example.
>
> Lionel
>
> Looks great!  Is it possible to change the size/position of the legend -
> the fonts certainly seem too big, and the whole box overlaps the figure.
>
> Derek
Hi Derek,
yes it's possible, with:
freq,ax=3Dwindrose.windplot(vent['VV'],vent['DV'],counts=3DFalse,speed_clas=
ses=3D[1,2,3,4,5],sectors=3D16,style=3D'bar')
draw()
fig=3Dax.get_figure()
for leg in fig.legends:
	setp(leg.get_texts(), fontsize=3D8)

>
>
>
>
> -------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job
> easier Download IBM WebSphere Application Server v.1.0.1 based on Apache
> Geronimo
> http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D120709&bid=3D263057&dat=
=3D121642
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

=2D-=20
Lionel Roubeyrie - lro...@li...
LIMAIR
http://www.limair.asso.fr

Lionel Roubeyrie <lroubeyrie@...> writes:

> 
> Hi all,
> continuing on the windroses, here the third version of windrose.py, with two 
> others styles (line and bar2 (is it what you want Derek?)). I want to know 
> how it's possible to modify the legends to be on axes, and not on the figure 
> like in the ex4.png subplots example. 

Lionel

Looks great!  Is it possible to change the size/position of the legend - the
fonts certainly seem too big, and the whole box overlaps the figure.

Derek

Hi, everyone:
is there any function in any libraray to do a multidimensional downhill
minimization for a given function?
Thanks

On Thu, Oct 12, 2006 at 07:00:05PM -0500, John Hunter wrote:
> >>>>> "Stefan" =3D=3D Stefan van der Walt <st...@su...> writes:
>     Stefan>             if hasattr(val, 'size') and val.size > 6:
>=20
> This looks like a numpy vs numeric thing here so we want to make sure
> the solution works across packages.  This is an attempt to pretty
> print an array that is long by snipping it.

The problem slips in when you have a scalar, which has attribute
'shape' but no length.  The above line works for Numeric, but I don't
have numarray to test on.

>     Stefan> Now on to my real question.  Say I have a screen of 4x4
>     Stefan> subplots, created with subplot(221).  I'd like to be able
>     Stefan> to split the screen as such: __ __
>     Stefan> |  |  |
>     Stefan> |  |  |
>     Stefan> |--| |
>     Stefan> |  |  |
>     Stefan> |__|__|
>=20
>     Stefan> I.e., to draw in the window on the right:
>=20
> ax1 =3D subplot(221)
> ax2 =3D subplot(223)
> ax3 =3D subplot(122)

That is so simple I never even expected it to work!  Thanks a lot.

Cheers
St=E9fan

>>>>> "Stefan" == Stefan van der Walt <st...@su...> writes:

    Stefan> One way to fix this would be to change line 379 in
    Stefan> artist.py from

    Stefan>             if hasattr(val, 'shape') and len(val)>6:
    Stefan> to

    Stefan>             if hasattr(val, 'size') and val.size > 6:

This looks like a numpy vs numeric thing here so we want to make sure
the solution works across packages.  This is an attempt to pretty
print an array that is long by snipping it.

    Stefan> Now on to my real question.  Say I have a screen of 4x4
    Stefan> subplots, created with subplot(221).  I'd like to be able
    Stefan> to split the screen as such: __ __
    Stefan> |  |  |
    Stefan> |  |  |
    Stefan> |--| |
    Stefan> |  |  |
    Stefan> |__|__|

    Stefan> I.e., to draw in the window on the right:

ax1 = subplot(221)
ax2 = subplot(223)
ax3 = subplot(122)

should do it.

JDH