matplotlib-users Mailing List for matplotlib (Page 3)

Thanks very much, guys!

-Gary


----- Original Message ----- 
From: "Todd Miller" <jm...@st...>
To: "John Hunter" <jdh...@ni...>
Cc: "Gary Pajer" <pa...@in...>; <mat...@li...>
Sent: Wednesday, March 17, 2004 11:38 AM
Subject: Re: [Matplotlib-users] TkAgg in a Tkinter GUI frame?


> On Tue, 2004-03-16 at 07:29, John Hunter wrote:
> > >>>>> "Gary" == Gary Pajer <pa...@in...> writes:
> >
> >     Gary> I've been poking aroung a bit trying to see if I can locate
> >     Gary> my plots in a frame in my own Tkinter GUI rather than a
> >     Gary> stand-alone window.
> >
> >     Gary> It looks like it might be possible if I can, for example,
> >     Gary> extract the frame instance from a TkAgg plot (or something
> >     Gary> like that)
> >
> >     Gary> Can anyone comment or suggest a route?
> >
> > Todd, is this easily doable?
>
> As we all hoped, this was easy and is now in CVS.  I refactored
> FigureManagerTkAgg and added FigureCanvasTkAgg to support the new
> functionality.  FigureCanvasTkAgg defines get_tk_widget() to return
> whatever tk widget is used to implement it. Currently it returns a Tk
> canvas but I thought it would be best not to assume that will always be
> the case.   Like other Tk widgets,  FigureCanvasTkAgg wants to be told
> who its parent widget (master) is at construction time.
>
> > Could you write an embedding_in_tk.py
> > example along the lines of embedding_in_gtk and embedding_in_wx?
>
> This is also in CVS.
>
> >
> Regards,
> Todd
>
> >
> >
> > -------------------------------------------------------
> > This SF.Net email is sponsored by: IBM Linux Tutorials
> > Free Linux tutorial presented by Daniel Robbins, President and CEO of
> > GenToo technologies. Learn everything from fundamentals to system
> > administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
> > _______________________________________________
> > Matplotlib-users mailing list
> > Mat...@li...
> > https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> -- 
> Todd Miller <jm...@st...>
>

On Wed, 2004-03-17 at 11:24, John Hunter wrote:
> >>>>> "Todd" == Todd Miller <jm...@st...> writes:
> 
>     Todd> This is also in CVS.
> 
> Nice work!
> 
> In a quick test, the only thing I noticed is that the figure window
> does not resize when embedded but does on a normal plot.  I don't know
> if this is a bug or a feature :-). 

I implemented this since you brought it up and I think the new way
qualifies as a feature.  The widget packing can be used to drive either
kind of behavior: resize-able or not-resize-able.  The demo now resizes.

Regards,
Todd

>  In backend_gtk, I handle the
> configure event in the FigureCanvas.
> 
> Gary, if you need a src snapshot (the regular CVS mirrors can lag
> sometimes by a couple of days), let me know.
> 
> JDH
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
-- 
Todd Miller <jm...@st...>

>>>>> "Todd" == Todd Miller <jm...@st...> writes:

    Todd> This is also in CVS.

Nice work!

In a quick test, the only thing I noticed is that the figure window
does not resize when embedded but does on a normal plot.  I don't know
if this is a bug or a feature :-).  In backend_gtk, I handle the
configure event in the FigureCanvas.

Gary, if you need a src snapshot (the regular CVS mirrors can lag
sometimes by a couple of days), let me know.

JDH

On Tue, 2004-03-16 at 07:29, John Hunter wrote:
> >>>>> "Gary" == Gary Pajer <pa...@in...> writes:
> 
>     Gary> I've been poking aroung a bit trying to see if I can locate
>     Gary> my plots in a frame in my own Tkinter GUI rather than a
>     Gary> stand-alone window.
> 
>     Gary> It looks like it might be possible if I can, for example,
>     Gary> extract the frame instance from a TkAgg plot (or something
>     Gary> like that)
> 
>     Gary> Can anyone comment or suggest a route?
> 
> Todd, is this easily doable?  

As we all hoped, this was easy and is now in CVS.  I refactored
FigureManagerTkAgg and added FigureCanvasTkAgg to support the new
functionality.  FigureCanvasTkAgg defines get_tk_widget() to return
whatever tk widget is used to implement it. Currently it returns a Tk
canvas but I thought it would be best not to assume that will always be
the case.   Like other Tk widgets,  FigureCanvasTkAgg wants to be told
who its parent widget (master) is at construction time.  

> Could you write an embedding_in_tk.py
> example along the lines of embedding_in_gtk and embedding_in_wx?

This is also in CVS.

> 
Regards,
Todd

> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
-- 
Todd Miller <jm...@st...>

It worked gret

Thanks

Jean-Baptiste

On Wed, 17 Mar 2004 18:33:52 +0800
LUK ShunTim <shu...@po...> wrote:

> Jean-Baptiste Cazier wrote:
> > S=E6l !
> >=20
> >=20
> > I am trying to keep up  with the versions of matplotlib by sucessively =
installing them.=20
> > However I have to install it on many workstations that share folders ov=
er the network
> >=20
> > I would like to be able to install matplotlib into such a folder to avo=
id installing it every time on each machine, but only once
> > It works fine with Numeric and other packages, but not with matplotlib:
> >=20
> >  - I install matplotlib on my own workstation.
> >  - I copied recursively the folder  /usr/lib/python2.2/site-packages/ma=
tplotlib into /home/stat/Pygtk/matplotlib
> >  - And then in my program I have added the location of the library to m=
y path:
> > sys.path.append("/home/stat/Pygtk/matplotlib")
> >=20
> > But I get the error message
> > [lws11] ~/devel/novi $  novi.py -m map  mixed.test.header.out
> > Traceback (most recent call last):
> >   File "./novi.py", line 19, in ?
> >     import matplotlib.matlab as Plot
> >   File "/home/stat/Pygtk/matplotlib/__init__.py", line 225, in ?
> >     defaultParams =3D {
> >   File "/home/stat/Pygtk/matplotlib/__init__.py", line 154, in get_data=
_path
> >     raise RuntimeError('Could not find the matplotlib data files')
> > RuntimeError: Could not find the matplotlib data files
> >=20
> >=20
> > Which other folder do I need to install or set a path to ?
> >=20
> > Thanks
> >=20
> > Kve=F0ja
> >=20
> > Jean-Baptiste
> >=20
>=20
> Hello,
>=20
> You may have to set the MATPLOTLIBDATA environmental variable to where=20
> you store the data (fonts etc) if you install in a non-default location.
>=20
> Regards,
> ST


--=20
-----------------------------
Jea...@de...

Department of Statistics
deCODE genetics     Sturlugata,8
570 2993          101 Reykjav=EDk

Jean-Baptiste Cazier wrote:
> Sæl !
> 
> 
> I am trying to keep up  with the versions of matplotlib by sucessively installing them. 
> However I have to install it on many workstations that share folders over the network
> 
> I would like to be able to install matplotlib into such a folder to avoid installing it every time on each machine, but only once
> It works fine with Numeric and other packages, but not with matplotlib:
> 
>  - I install matplotlib on my own workstation.
>  - I copied recursively the folder  /usr/lib/python2.2/site-packages/matplotlib into /home/stat/Pygtk/matplotlib
>  - And then in my program I have added the location of the library to my path:
> sys.path.append("/home/stat/Pygtk/matplotlib")
> 
> But I get the error message
> [lws11] ~/devel/novi $  novi.py -m map  mixed.test.header.out
> Traceback (most recent call last):
>   File "./novi.py", line 19, in ?
>     import matplotlib.matlab as Plot
>   File "/home/stat/Pygtk/matplotlib/__init__.py", line 225, in ?
>     defaultParams = {
>   File "/home/stat/Pygtk/matplotlib/__init__.py", line 154, in get_data_path
>     raise RuntimeError('Could not find the matplotlib data files')
> RuntimeError: Could not find the matplotlib data files
> 
> 
> Which other folder do I need to install or set a path to ?
> 
> Thanks
> 
> Kveðja
> 
> Jean-Baptiste
> 

Hello,

You may have to set the MATPLOTLIBDATA environmental variable to where 
you store the data (fonts etc) if you install in a non-default location.

Regards,
ST

S=E6l !


I am trying to keep up  with the versions of matplotlib by sucessively inst=
alling them.=20
However I have to install it on many workstations that share folders over t=
he network

I would like to be able to install matplotlib into such a folder to avoid i=
nstalling it every time on each machine, but only once
It works fine with Numeric and other packages, but not with matplotlib:

 - I install matplotlib on my own workstation.
 - I copied recursively the folder  /usr/lib/python2.2/site-packages/matplo=
tlib into /home/stat/Pygtk/matplotlib
 - And then in my program I have added the location of the library to my pa=
th:
sys.path.append("/home/stat/Pygtk/matplotlib")

But I get the error message
[lws11] ~/devel/novi $  novi.py -m map  mixed.test.header.out
Traceback (most recent call last):
  File "./novi.py", line 19, in ?
    import matplotlib.matlab as Plot
  File "/home/stat/Pygtk/matplotlib/__init__.py", line 225, in ?
    defaultParams =3D {
  File "/home/stat/Pygtk/matplotlib/__init__.py", line 154, in get_data_path
    raise RuntimeError('Could not find the matplotlib data files')
RuntimeError: Could not find the matplotlib data files


Which other folder do I need to install or set a path to ?

Thanks

Kve=F0ja

Jean-Baptiste

--=20
-----------------------------
Jea...@de...

Department of Statistics
deCODE genetics     Sturlugata,8
570 2993          101 Reykjav=EDk

Hello,
I sent a message to the numarray website to let them know that
when they say "numarray lacks a plotting module" that's no longer
true since matplotlib now works with numarray.

However, when I looked at the matplotlib install page
http://matplotlib.sourceforge.net/installing.html
it currently says
"matplotlib requires python 2.2+, Numeric."
and further down
"Note, you'll want the "numpy", not "numarray" files."

I'm assuming the matplotlib support for numarray means that Numeric
is no longer a requirement, is this true? If so it looks like the
install page needs to be updated to avoid confusing numarray users.

Regards
Steve

On Tue, 2004-03-16 at 07:29, John Hunter wrote:
> >>>>> "Gary" == Gary Pajer <pa...@in...> writes:
> 
>     Gary> I've been poking aroung a bit trying to see if I can locate
>     Gary> my plots in a frame in my own Tkinter GUI rather than a
>     Gary> stand-alone window.
> 
>     Gary> It looks like it might be possible if I can, for example,
>     Gary> extract the frame instance from a TkAgg plot (or something
>     Gary> like that)
> 
>     Gary> Can anyone comment or suggest a route?
> 

First an apology.  I didn't see this until late yesterday when Perry
alerted me.  I wasn't subscribed to matplotlib-users.

> Todd, is this easily doable?  

Frankly, I'm not sure.  I'll try to work it out this afternoon.  I
imagine if it's possible in GTK it's possible in Tk.

> Could you write an embedding_in_tk.py
> example along the lines of embedding_in_gtk and embedding_in_wx?

I'll try and let you know.

> 
> The FigureCanvas derived instance in backend_* is meant to be a widget
> you could plug into a GUI, but I notice you took a different tack in
> backend_tkagg.  The idea behind this is that users may want to embed a
> figure as part of a GUI window, eg in a widget container, without
> necessarily bringing the toolbar along or using the whole window.
> Would this be an easy refactoring of backend_tkagg?

We'll see.

Regards,
Todd

-- 
Todd Miller <jm...@st...>

What's new in matplotlib-0.52

Image support

  Basic image support. Images can be specified by Numeric float arrays

     imshow(X)
     If X is MxN, assume luminance (grayscale)
     If X is MxNx3, assume RGB
     If X is MxNx4, assume RGBA

     imshow(X, cmap)  # plot X using colormap; see examples/pcolor_demo2.py

  see help(imshow) and the image_demo*.py examples in the matplotlib
  src distribution. Set BUILD_IMAGE in setup.py for image
  support. Currently available on Agg, GTKAgg, TkAgg and GTK
  backends. win32 GTK users should use GTKAgg unless your pygtk is
  compiled with Numeric support.

  The pseudocolor images generated with imshow are 8 million times
  faster than pcolors.

Figure legends

  In addition to adding legends to the axes with the legend command,
  you can place legends anywhere in the figure with figlegend

fill command

  Andrew Straw wrote a fill command to plot filled polygons. See
  fill_demo.py

Make 2D spectrograms with specgram. 

  Requires image support; see specgram_demo.py 

Bugfixes and minor improvements

    * Tk : Fixed a close figure bug in interactive mode

    * GTK : Much improved mathtext performance thanks to patch by
      Trevor Blackwell

    * All : Fixed a bug that showed up in successive calls to plot
      with just one plot argument

Downloads at
https://sourceforge.net/project/showfiles.php?group_id=80706&package_id=82474&release_id=224080

>>>>> "matthew" == matthew arnison <ma...@ca...> writes:

    matthew> I can't seem to get multiple plots to draw on the same
    matthew> figure under TkAgg.  In the below example, the cos()
    matthew> doesn't show up, and the tan produces an error.  This
    matthew> seems to be repeatable for arbitrary cases of trying to
    matthew> plot 2 or 3 lines together on the one plot.

Found and fixed.  This crept into the functions that process plot
arguments when I added matplotlibrc support there.  In axes.py, search
for def _plot_1_arg and replace the first part of it with

    def _plot_1_arg(self, y, **kwargs):
        if self.count==0:
            color = self.firstColor
        else:
            color = self.colors[int(self.count % self.Ncolors)]
                      ^^^ this is the critical part

    matthew> Cheers and thanks for matplotlib, Matthew.

Thanks for the bug report!

JDH

>>>>> "Gary" == Gary Pajer <pa...@in...> writes:

    Gary> I've been poking aroung a bit trying to see if I can locate
    Gary> my plots in a frame in my own Tkinter GUI rather than a
    Gary> stand-alone window.

    Gary> It looks like it might be possible if I can, for example,
    Gary> extract the frame instance from a TkAgg plot (or something
    Gary> like that)

    Gary> Can anyone comment or suggest a route?

Todd, is this easily doable?  Could you write an embedding_in_tk.py
example along the lines of embedding_in_gtk and embedding_in_wx?

The FigureCanvas derived instance in backend_* is meant to be a widget
you could plug into a GUI, but I notice you took a different tack in
backend_tkagg.  The idea behind this is that users may want to embed a
figure as part of a GUI window, eg in a widget container, without
necessarily bringing the toolbar along or using the whole window.
Would this be an easy refactoring of backend_tkagg?

Thanks,
JDH

I can't seem to get multiple plots to draw on the same figure under TkAgg.  
In the below example, the cos() doesn't show up, and the tan produces an
error.  This seems to be repeatable for arbitrary cases of trying to plot
2 or 3 lines together on the one plot.

Cheers and thanks for matplotlib,
Matthew.

$ python
Python 2.2.2 (#1, Feb 24 2003, 19:13:11)
[GCC 3.2.2 20030222 (Red Hat Linux 3.2.2-4)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> from matplotlib.matlab import *
>>> xx = arange(0, 3, 0.1)
>>> plot(sin(xx))
[<matplotlib.lines.Line2D instance at 0x819613c>]
>>> plot(cos(xx))
[<matplotlib.lines.Line2D instance at 0x8636444>]
>>> plot(tan(xx))
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.2/site-packages/matplotlib/matlab.py", line 788, 
in plot
    draw_if_interactive()
  File 
"/usr/lib/python2.2/site-packages/matplotlib/backends/backend_tkagg.py",
line 46, in draw_if_interactive
    figManager.show()
  File 
"/usr/lib/python2.2/site-packages/matplotlib/backends/backend_tkagg.py",
line 127, in show
    self.canvas.draw()
  File 
"/usr/lib/python2.2/site-packages/matplotlib/backends/backend_agg.py", 
line 334, in draw
    self.figure.draw(self.renderer)
  File "/usr/lib/python2.2/site-packages/matplotlib/artist.py", line 79, 
in draw    self._draw(renderer, *args, **kwargs)
  File "/usr/lib/python2.2/site-packages/matplotlib/figure.py", line 81, 
in _draw
    for a in self.axes: a.draw(renderer)
  File "/usr/lib/python2.2/site-packages/matplotlib/artist.py", line 79, 
in draw    self._draw(renderer, *args, **kwargs)
  File "/usr/lib/python2.2/site-packages/matplotlib/axes.py", line 493, in 
_draw    line.draw(renderer)
  File "/usr/lib/python2.2/site-packages/matplotlib/artist.py", line 79, 
in draw    self._draw(renderer, *args, **kwargs)
  File "/usr/lib/python2.2/site-packages/matplotlib/lines.py", line 177, 
in _draw
    gc.set_foreground(self._color)
  File "/usr/lib/python2.2/site-packages/matplotlib/backend_bases.py", 
line 230, in set_foreground
    self._rgb = colorConverter.to_rgb(fg)
  File "/usr/lib/python2.2/site-packages/matplotlib/colors.py", line 82, 
in to_rgb
    error_msg(msg)
NameError: global name 'error_msg' is not defined
>>>

It workls great now !


Thanks

Jean-Baptiste

On Sun, 14 Mar 2004 13:26:20 -0600
John Hunter <jdh...@ni...> wrote:

> >>>>> "Jean-Baptiste" =3D=3D Jean-Baptiste Cazier <Jean-Baptiste.cazier@d=
ecode.is> writes:
>=20
>   lh=3Dlen(h.haplist)
>   Xrange=3D[map[m] for m in [y for x,y in h.haplist]]
>   Yrange=3D[-math.log10(h.pval) for i in range(lh)]
>   line=3Dcanvas.figure.axes[0].plot(Xrange, Yrange, symbol)
>   line.set_visible(Plot.False)
>=20
>     Jean-Baptiste> Any idea ?
>=20
> plot always return *a list* of lines.  So you need to do
>=20
>   lines =3D canvas.figure.axes[0].plot(Xrange, Yrange, symbol)
>   lines[0].set_visible(Plot.False) # get the first element of the list
>=20
> JDH


--=20
-----------------------------
Jea...@de...

Department of Statistics
deCODE genetics     Sturlugata,8
570 2993          101 Reykjav=EDk

>>>>> "Jean-Baptiste" == Jean-Baptiste Cazier <Jea...@de...> writes:

  lh=len(h.haplist)
  Xrange=[map[m] for m in [y for x,y in h.haplist]]
  Yrange=[-math.log10(h.pval) for i in range(lh)]
  line=canvas.figure.axes[0].plot(Xrange, Yrange, symbol)
  line.set_visible(Plot.False)

    Jean-Baptiste> Any idea ?

plot always return *a list* of lines.  So you need to do

  lines = canvas.figure.axes[0].plot(Xrange, Yrange, symbol)
  lines[0].set_visible(Plot.False) # get the first element of the list

JDH

S=E6l !


Thanks for the change of the artist class. It should do exaclty what I want
I modified artist accordingly (no more 'transform' defined ?
I can run you example witout problem.

But If I implement it in my program which is embedded within gtk with the o=
bject_picker, it does not work:

If I try to set_visible on lines I just define with plot I get an error mes=
sage
Traceback (most recent call last):
  File "./novi.py", line 604, in ?
    main()
  File "./novi.py", line 576, in main
    MaxP =3D plot_Haps(canvas, map, Haps)
  File "./novi.py", line 367, in plot_Haps
    line =3D plot_hap(canvas, map, h, sym[s])
  File "./novi.py", line 389, in plot_hap
    line.set_visible(Plot.False)
AttributeError: 'list' object has no attribute 'set_visible'

And this happens even if I redo the full installation of matplotlib 0.51:
...
not copying matplotlib/__init__.py (output up-to-date)
not copying matplotlib/_matlab_helpers.py (output up-to-date)
not copying matplotlib/afm.py (output up-to-date)
copying matplotlib/artist.py -> build/lib/matplotlib
not copying matplotlib/axes.py (output up-to-date)
not copying matplotlib/axis.py (output up-to-date)
...
running install_lib
not copying build/lib/matplotlib/__init__.py (output up-to-date)
not copying build/lib/matplotlib/_matlab_helpers.py (output up-to-date)
not copying build/lib/matplotlib/afm.py (output up-to-date)
copying build/lib/matplotlib/artist.py -> /usr/lib/python2.2/site-packages/=
matplotlib
not copying build/lib/matplotlib/axes.py (output up-to-date)
not copying build/lib/matplotlib/axis.py (output up-to-date)
not copying build/lib/matplotlib/backend_bases.py (output up-to
....
skipping byte-compilation of /usr/lib/python2.2/site-packages/matplotlib/_m=
atlab_helpers.py to
_matlab_helpers.pyc
skipping byte-compilation of /usr/lib/python2.2/site-packages/matplotlib/af=
m.py to afm.pyc
byte-compiling /usr/lib/python2.2/site-packages/matplotlib/artist.py to art=
ist.pyc
skipping byte-compilation of /usr/lib/python2.2/site-packages/matplotlib/ax=
es.py to axes.pyc
skipping byte-compilation of /usr/lib/python2.2/site-packages/matplotlib/ax=
is.py to axis.pyc
...


This is my routine:

def plot_hap(canvas, map, h , symbol, size=3DNone):
  """ plot the specific haplotype """

  lh=3Dlen(h.haplist)
  Xrange=3D[map[m] for m in [y for x,y in h.haplist]]
  Yrange=3D[-math.log10(h.pval) for i in range(lh)]
  line=3Dcanvas.figure.axes[0].plot(Xrange, Yrange, symbol)
  line.set_visible(Plot.False)
  return line


Any idea ?

Thanks

Jean-Baptiste


On Fri, 12 Mar 2004 12:29:18 -0600
John Hunter <jdh...@ni...> wrote:

> >>>>> "Jean-Baptiste" =3D=3D Jean-Baptiste Cazier <Jean-Baptiste.cazier@d=
ecode.is> writes:
>=20
>     Jean-Baptiste> S=E6l !  Is there a way to "turn-off" lines without
>     Jean-Baptiste> removing the data ?  My goal is to hide some lines
>     Jean-Baptiste> in a plot wihtout losing the data so I can show it
>     Jean-Baptiste> again later.  I can do # Hide the line
>     Jean-Baptiste> x=3Dline.get_xdata() y=3Dline.get_ydata()
>     Jean-Baptiste> line.set_data([],[])
>=20
> This can be done very easily (for any artist) with a minor
> modification of artist.py.  The base class forewards all drawing to
> the derived classes so no other changes are required.  Just replace
> artist.py with the attached file below and then you can do:
>=20
>     from matplotlib.matlab import *
>=20
>     x =3D arange(0.0, 1.0, 0.05)
>     l1, l2 =3D plot(x, sin(2*pi*x), x, sin(4*pi*x))
>     l1.set_visible(False)
>     show()
>=20
>=20


--=20
-----------------------------
Jea...@de...

Department of Statistics
deCODE genetics     Sturlugata,8
570 2993          101 Reykjav=EDk

I've been poking aroung a bit trying to see if I can locate my plots in a
frame in my own Tkinter GUI rather than a stand-alone window.

It looks like it might be possible if I can, for example, extract the frame
instance from a TkAgg plot (or something like that)

Can anyone comment or suggest a route?

-gary

>>>>> "Jean-Baptiste" =3D=3D Jean-Baptiste Cazier <Jean-Baptiste.cazier@d=
ecode.is> writes:

    Jean-Baptiste> S=E6l !  Is there a way to "turn-off" lines without
    Jean-Baptiste> removing the data ?  My goal is to hide some lines
    Jean-Baptiste> in a plot wihtout losing the data so I can show it
    Jean-Baptiste> again later.  I can do # Hide the line
    Jean-Baptiste> x=3Dline.get_xdata() y=3Dline.get_ydata()
    Jean-Baptiste> line.set_data([],[])

This can be done very easily (for any artist) with a minor
modification of artist.py.  The base class forewards all drawing to
the derived classes so no other changes are required.  Just replace
artist.py with the attached file below and then you can do:

    from matplotlib.matlab import *

    x =3D arange(0.0, 1.0, 0.05)
    l1, l2 =3D plot(x, sin(2*pi*x), x, sin(4*pi*x))
    l1.set_visible(False)
    show()

S=E6l !


Is there a way to "turn-off" lines without removing the data ?
My goal is to hide some lines in a plot wihtout losing the data so I can sh=
ow it again later.
I can do=20
# Hide the line
x=3Dline.get_xdata()
y=3Dline.get_ydata()
line.set_data([],[])

# Reset the line
line.set_data(x,y)

But I would prefer I more elegant way like
line.hide()
line.show()

Would it be possible to get something like that ?

Thanks

Jean-Baptiste


--=20
-----------------------------
Jea...@de...

Department of Statistics
deCODE genetics     Sturlugata,8
570 2993          101 Reykjav=EDk

In order to "freeze" your matplotlib-0.51 application using the Agg
backend into a standalone executable, the following are the steps you
should take:

1. Write your setup.py file that py2exe will use for compiling. Example:

    # setup.py
    from distutils.core import setup
    import py2exe
    import glob
	
    data = glob.glob(r'C:\Python23\share\matplotlib\*')

    setup(
        console=["your_file.py"],
        data_files=[("share",data)]
    )

. Note that we use "data" here to copy the matplotlibdata "share" folder
contents into your /dist/share directory.

2. Insert the following import statements into your_file.py, which is
the file containing your matplotlib code:

    import matplotlib
    import matplotlib.ft2font
    import matplotlib.backends.backend_agg
    import ttfquery
    from matplotlib.matlab import *
    
3. Now we must take care of fonts. We're assuming that the machine we're
installing to doesn't have the MATPLOTLIBDATA or TTFPATH environment
variables set, so we must do the following:

    In ttfquery\_scriptregistry.py comment out these lines:

    ### more robust registry-file location by John Hunter...
    #if os.environ.has_key('HOME'):
    #    registryFile = os.path.join( os.environ['HOME'], ".font.cache")
    #else:
	# OpenGLContext uses the Application Data directory for win32,
	# should consider porting that code here...
      #    registryFile = os.path.join( os.path.split(__file__)[0],
"font.cache")

    And replace with:

    registryFile = r'share\.font.cache'

    In matplotlib\__init__.py comment out these lines in the
get_data_path() method:

   	#path = os.path.join(distutils.sysconfig.PREFIX, 'share',
'matplotlib')
        #if os.path.isdir(path): return path

            #path =
os.path.join(os.sep.join(__file__.split(os.sep)[:-5]), 
    #                    'share','matplotlib')
    #if os.path.isdir(path): return path

    #if os.environ.has_key('MATPLOTLIBDATA'):
    #    path = os.environ['MATPLOTLIBDATA']
    #    if os.path.isdir(path): return path
    #raise RuntimeError('Could not find the matplotlib data files')

    And replace with:

    return 'share'

    Also, while you have __init__.py open, in defaultParams you should
change the default backend to 'Agg' and         
    default fontname to 'Vera' since it ships with matplotlib. 

4. Copy your .font.cache file from /ttfquery into your matplotlibdata
/share directory. Your system did a system scan of fonts to generate
this file the first time that Agg was run but your "clean" system won't
have it. By copying it to share, py2exe will include it in the generated
/dist/share directory. 


That should be it. Many thanks to John Hunter for helping me hack
through this, and I hope this information will find its way to others so
it can save them lots of time. 

--Matt

John I foud it with the "print trick" thank you!!!

the main proble was the self.fig  which I replaced with the self.canvas
and it is now working,

Thanks again

On Thu, 2004-03-11 at 13:49, Flavio Codeco Coelho wrote:

> -----Forwarded Message-----
> From: Flavio Codeco Coelho <fcc...@fi...>
> To: mat...@li...
> <mat...@li...>
> Subject: [Matplotlib-users] embedding in wx segfault
> Date: Thu, 11 Mar 2004 12:53:40 -0300
> 
> 
> 
> Hi I have a variant of the embedding_in_wx script that worked fine on
> my previous version of matplotlib. Now, with version 0.51 it's seg
> faulting. I updated the script to reflect the changes in the library
> but still... Can anyone see what may be going wrong?
> 
> thanks
> 
> the script is attached

-----Forwarded Message-----
From: Flavio Codeco Coelho <fcc...@fi...>
To: mat...@li...
<mat...@li...>
Subject: [Matplotlib-users] embedding in wx segfault
Date: Thu, 11 Mar 2004 12:53:40 -0300



Hi I have a variant of the embedding_in_wx script that worked fine on my
previous version of matplotlib. Now, with version 0.51 it's seg
faulting. I updated the script to reflect the changes in the library but
still... Can anyone see what may be going wrong?

thanks

the script is attached

>>>>> "Flavio" == Flavio Codeco Coelho <fcc...@fi...> writes:

    Flavio> Hi I have a variant of the embedding_in_wx script that
    Flavio> worked fine on my previous version of matplotlib. Now,
    Flavio> with version 0.51 it's seg faulting. I updated the script
    Flavio> to reflect the changes in the library but still... Can
    Flavio> anyone see what may be going wrong?

Please resend as an attachment - your lines are wrapped and hence the
code is doubly difficult to debug.

A couple of suggestions: wx swallows exceptions.  This makes it
virtually impossible to debug scripts using wx since you cannot read
the stack trace.  The problem may be a simple syntax error in your
script having nothing to do with matplotlib or the wx backend per se,
and you wouldn't see it because of the exception swallowing problem.

Workarounds: 

  1) insert some print statements at the top of each function so you
     can see when the failure occurs.  This is essentially what Jeremy
     does in backend_wx.  Change the _DEBUG level to a low number in
     backend_wx to get more debugging information.  

  2) try running pychecker on your script.

A lifetime supply of free bugfixes to any wx guru who knows how to fix
this problem - Jeremy and I have certainly suffered with it long enough.

JDH

Hi I have a variant of the embedding_in_wx script that worked fine on my
previous version of matplotlib. Now, with version 0.51 it's seg
faulting. I updated the script to reflect the changes in the library but
still... Can anyone see what may be going wrong?

thanks

here is my script:

-----------------------------------------------------------------------------
# Name:        PlotFigure.py
# Purpose:     Plotting frame that contains the plots generated by Model
Buider
#
# Author:      Flavio C. Coelho
#
# Created:     2004/09/01
# RCS-ID:      $Id: PlotFigure.py,v 1.1 2004/01/13 10:51:43 fccoelho Exp
$
# Copyright:   (c) 2003
# Licence:     GPL
# Obs:   This code was based on Jeremy Donoghue's embedding_in_wx.py
included with 
#       matplotlib.
#-----------------------------------------------------------------------------
#Boa:Frame:PlotFigure

import matplotlib
matplotlib.use('WX')

from matplotlib.backends.backend_wx import Toolbar, FigureCanvasWx,
FigureManager
from matplotlib.figure import Figure
from matplotlib.axes import Subplot
import matplotlib.numerix as numpy
from RandomArray import *

from wxPython.wx import *

def create(parent):
    return PlotFigure(parent)

[wxID_PLOTFIGURE] = map(lambda _init_ctrls: wxNewId(), range(1))

class PlotFigure(wxFrame):
    def _init_ctrls(self, prnt):
        # generated method, don't edit
        wxFrame.__init__(self, id=wxID_PLOTFIGURE, name='Output',
parent=prnt,
              pos=wxPoint(480, 335), size=wxSize(640, 480),
              style=wxDEFAULT_FRAME_STYLE, title='Results')
        self.SetClientSize(wxSize(1280, 893))

    def __init__(self, parent):
        self._init_ctrls(parent)
        self.fig = Figure((10,8), 75)
        self.canvas = FigureCanvasWx(self,-1, self.fig)
        self.toolbar = Toolbar(self.canvas)
        self.toolbar.Realize()

		# On Windows, default frame size behaviour is incorrect
        # you don't need this under Linux
        tw, th = self.toolbar.GetSizeTuple()
        fw, fh = self.canvas.GetSizeTuple()
        self.toolbar.SetSize(wxSize(fw, th))
        
        # Create a figure manager to manage things
        self.figmgr = FigureManager(self.canvas, 1, self)
        
        # Now put all into a sizer
        sizer = wxBoxSizer(wxVERTICAL)
		# This way of adding to sizer prevents resizing
        #sizer.Add(self.fig, 0, wxLEFT|wxTOP)
		
		# This way of adding to sizer allows resizing
        sizer.Add(self.fig, 1, wxLEFT|wxTOP|wxGROW)
		
		# Best to allow the toolbar to resize!
        sizer.Add(self.toolbar, 0, wxGROW)
        self.SetSizer(sizer)
        self.Fit()
        
    def plot_data(self, x,y):
        """
        This function will plot the time series as output by odeint.
        """
        # Use this line if using a toolbar
        a = self.figmgr.add_subplot(111)
        
        # Or this one if there is no toolbar
        #a = Subplot(self.fig, 111)
        
        nvar = min(y[0].shape)
        
        for i in range(nvar):
            a.plot (x,y[0][:,i])
        #plot (x,timeseries[1])
        a.set_xlabel('Time')
        a.set_ylabel('Y[i]')
        a.set_title('Time series')
#---generating tuple of
legends-------------------------------------------------
        b = range(nvar)
        leg = tuple(['y['+str(i)+']' for i in b])
#-------------------------------------------------------------------------------
        a.legend(leg)
        
        self.toolbar.update()
    def plotStats(self,x, ts):
        """
        This function will plot prior an posteriors for the model
variables and parameters.
        """
        # Use this line if using a toolbar
        a = self.figmgr.add_subplot(111)
        
        # Or this one if there is no toolbar
        #a = Subplot(self.fig, 111)
        
        
        a.plot(x,ts[0], 'r-o',x,ts[1],'b-.', x,ts[2],'b-.')
        
        
    def plotDist(data,vname):
        """
        Plots histograms ofthe prior and posterior distributions of the
model components
        meldOut is the output of the Melding.SIR function:
(w,qtiltheta,qtilphi,q1est)
        """
        # Use this line if using a toolbar
        a = self.figmgr.add_subplot(111)
        
        # Or this one if there is no toolbar
        #a = Subplot(self.fig, 111)
        
        nb, bins, patches = a.hist(data, normed=1)
        a.set_title('Prior Distribution')
        a.set_xtitle(vname)
        
        nvp = len(priors) # Get number of variable + parameters in the
model)
        nlik = len (liks) # Get number of likelihood functions
        
        for i in range(nvp):
            nb, bins, patches = a.hist(priors[i], normed=1)
            if i > 0:
                a.figure(i+1)
            a.set_title('Prior Distribution')
     
            
        
        

    def GetToolBar(self):
        # You will need to override GetToolBar if you are using an 
        # unmanaged toolbar in your frame
        return self.toolbar
        
if __name__ == '__main__':
    app = wxPySimpleApp()
    frame = create(None)
    x = normal(0,1,50)
    y = normal(0,1,(5,50))
    frame.plot_data(x,y)
    frame.Show()
    app.MainLoop()

>>>>> "Kuzminski," == Kuzminski, Stefan R <SKu...@fa...> writes:

    Kuzminski,> Is there a way for the legend to be outside of the
    Kuzminski,> axis?  Perhaps rendered separately.

The current implementation of legend takes an axes instance in it's
constructor.  I just did a read through of that code and it would be
trivial to factor it out.  The only thing the legend uses that
instance for is the background color!

You could have a figure legend much in the same way you have figure
text.  I can probably do it in 20 minutes.   When I get 20 minutes,
I'll send you a patch....

JDH