matplotlib-users Mailing List for matplotlib

On 9/1/06, Kenny Ortmann <ya...@em...> wrote:
> hey guys i got the subversion from the site and I am trying to install it
> on windows.
>
> I changed dir into the matplotlib dir that includes the setup.py file.
>
> run python setup.py install, and im getting a wierd error.  i left the
> topmost lines along with the error.  has anyone seen anything like this
> before?
>
> building 'matplotlib.enthought.traits.ctraits' extension
> creating build\temp.win32-2.4\Release\lib
> creating build\temp.win32-2.4\Release\lib\matplotlib
> creating build\temp.win32-2.4\Release\lib\matplotlib\enthought
> creating build\temp.win32-2.4\Release\lib\matplotlib\enthought\traits
> C:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\bin\cl.exe /c
> /nologo /Ox
>  /MD /W3 /GX /DNDEBUG -Ic:\Python24\include -Ic:\Python24\PC
> /Tclib/matplotlib/e
> nthought/traits/ctraits.c
> /Fobuild\temp.win32-2.4\Release\lib/matplotlib/enthoug
> ht/traits/ctraits.obj
> C:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\bin\link.exe /DLL
> /nologo
>  /INCREMENTAL:NO /LIBPATH:c:\Python24\libs /LIBPATH:c:\Python24\PCBuild
> /EXPORT:
> initctraits
> build\temp.win32-2.4\Release\lib/matplotlib/enthought/traits/ctraits
> .obj /OUT:build\lib.win32-2.4\matplotlib\enthought\traits\ctraits.pyd
> /IMPLIB:bu
> ild\temp.win32-2.4\Release\lib/matplotlib/enthought/traits\ctraits.lib
> building 'matplotlib.backends._tkagg' extension
> C:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\bin\cl.exe /c
> /nologo /Ox
>  /MD /W3 /GX /DNDEBUG -Iwin32_static/include/tcl84 -I. -Isrc -Iswig
> -Iagg23/incl
> ude -I. -I. -Iwin32_static/include/tcl84\freetype2 -I.\freetype2
> -Isrc\freetype2
>  -Iswig\freetype2 -Iagg23/include\freetype2 -I.\freetype2 -I.\freetype2
> -Ic:\Pyt
> hon24\include -Ic:\Python24\PC /Tpsrc/_tkagg.cpp
> /Fobuild\temp.win32-2.4\Release
> \src/_tkagg.obj
> _tkagg.cpp
> src\_tkagg.cpp(28) : fatal error C1083: Cannot open include file: 'tk.h':
> No suc
> h file or directory
> error: Command ""C:\Program Files\Microsoft Visual Studio .NET
> 2003\Vc7\bin\cl.e
> xe" /c /nologo /Ox /MD /W3 /GX /DNDEBUG -Iwin32_static/include/tcl84 -I.
> -Isrc -
> Iswig -Iagg23/include -I. -I. -Iwin32_static/include/tcl84\freetype2
> -I.\freetyp
> e2 -Isrc\freetype2 -Iswig\freetype2 -Iagg23/include\freetype2
> -I.\freetype2 -I.\
> freetype2 -Ic:\Python24\include -Ic:\Python24\PC /Tpsrc/_tkagg.cpp
> /Fobuild\temp
> .win32-2.4\Release\src/_tkagg.obj" failed with exit status 2

You need to install the tcl/tk headers as Darren mentioned.  I just
install ActiveTcl and the build should pick up on it no problem.

For the benefit of the mailing list & developers, this fix works fine
for me (OS X Tiger 10.4.7)

Cheers John.

Dave

On 04/09/06, John Hunter <jdh...@ac...> wrote:
> >>>>> "David" == David Andrews <irb...@gm...> writes:
>
>     David> Hiya, i'm having a problem getting a valid postscript file
>     David> produced on OS X.  Using the example subplot_demo.py,
>     David> modified to include the line:
>
>     David>     savefig('subplot_demo')
>
>     David> in the place of the show() command, and running it using:
>
>     David>     python subplot_demo.py -dPS
>
>     David> Produces the .ps file (available here for reference:
>     David> http://www.irbdavid.com/misc/subplot_demo.ps ), which can't
>     David> be opened by Preview on os x - says its can't convert it to
>     David> PDF & cant open the file.  Checked the permissions etc on
>     David> the file and they're fine.
>
>     David> Its pretty much a default installation of matplotlib etc,
>     David> and output displays fine on screen, using the wxPython /
>     David> wxAgg thing (I think that's what i mean :D )
>
> We've noticed this on tiger several times.  Some default system fault
> is causing troubles though we haven't identified which one.  The
> recommended fix it to put the Vera fonts (which matplotlib ships with)
> first in your rc file (http://matplotlib.sf.net/matplotlibrc) and
> remove your ~/.matplotlib/ttf.cache before rerunning
>
> Ie, you fonts section of ex should look something like this:
>
>
>   font.serif          : Bitstream Vera Serif, New Century Schoolbook, Century Schoolbook L, Utopia, ITC Bookman, Bookman, Nimbus Roman No9 L, Times New Roman, Times, Palatino, Charter, serif
>   font.sans-serif     : Bitstream Vera Sans, Lucida Grande, Verdana, Geneva, Lucid, Arial, Helvetica, Avant Garde, sans-serif
>   font.cursive        : Apple Chancery, Textile, Zapf Chancery, Sand, cursive
>   font.fantasy        : Comic Sans MS, Chicago, Charcoal, Impact, Western, fantasy
>   font.monospace      : Bitstream Vera Sans Mono, Andale Mono, Nimbus Mono L, Courier New, Courier, Fixed, Terminal, monospace
>
> In future releases of mpl, this will be the default.
>
> JDH
>

>
>
> Message: 2
> Date: Mon, 04 Sep 2006 09:56:22 -0700
> From: Andrew Straw <str...@as...>
> Subject: Re: [Matplotlib-users] matplotlib, GTK and Threads
> To: Jon Roadley-Battin <jon...@gm...>,
>         mat...@li...
> Message-ID: <44F...@as...>
> Content-Type: text/plain; charset=ISO-8859-1
>
> Jon Roadley-Battin wrote:
> > The problem I have with with the non-GUI thread calling the "plot"
> > function.
>
> That's your problem -- you should only call the GUI code (including
> matplotlib, which eventually calls GTK) from a single thread. I suggest
> looking at the Queue module to pass your data to the GUI thread if it's
> not too much. If it's more, you could use threading.Lock() to regulate
> access to a giant array or something.
>
> -Andrew
>
>
>
> I was pretty convinced that is was due to the threading since early one
> with GTK I had some issues with it.
> I was hoping that there was some form of enable_threading method with
> matplotlib.
>
> The problem that makes this hard and what forced me down the thread route
> is that the thread polls the RS232 every second or so, thus it needs to be
> separate (ie threaded) from the GUI, this have been done with hte
> enter/leave thread statement for GTK.
>
> But now it seems that updating the Matplotlib MUST be done from within the
> GUI and not initilised by another thread (be it if the thread enabled GTK
> events) I must think of a way around it.
>
> The only other option is to start a GTK timer event (did use this
> originally, but they had problems with the whole asyncronous of the RS232)
> that every second check's if my plot array has been updated (via some flag)
> and if it has then update plot, this way it would be within the GUI-class
>
>
> I stipped the code down to its most basic instance (the window will have
> to be dragged wider)


is there any thread_init for matplotlib. short of having a timer within the
GUI class that checks if any new data is present I cant see a way around
this. It has to be done this way, the GUI doesn't instigate things, the GUI
only reports things, the thread gets the data at set intervals and then
updates the GUI

#!/usr/bin/env python
> import threading
> import time
> import math
> import sys
> import os
>
> import pygtk
> if sys.platform == 'win32':
>     os.environ['PATH'] += ';lib;'
> else:
>     pygtk.require('2.0')
> import gtk
> import gobject
> assert gtk.pygtk_version >= (1,99,16), 'pygtk should be >= 1.99.16'
> #import gtk.glade
>
>
> from pylab import *
> rcParams['numerix'] = 'numpy'
>
> import matplotlib
> matplotlib.use ('GTK')
> from matplotlib.figure import Figure
> from matplotlib.axes import Subplot
> from matplotlib.backends.backend_gtk import FigureCanvasGTK as
> FigureCanvas
>
>
> TIME = range(360)
> VOLT = [math.sin(math.radians(x)) for x in TIME]
> VOLT2 = [2]*360
> VOLT3 = [x for x in range(360)]
>
>
> class my_thread(threading.Thread):
>     def __init__(self,GUI):
>         super(my_thread, self).__init__()
>         self.GUI = GUI
>
>     def run(self):
>         time.sleep(10)
>         gtk.threads_enter()
>         try:
>             self.GUI.Graph([TIME,VOLT3])
>         finally:
>             gtk.threads_leave()
>
>
>
>
>
> gtk.gdk.threads_init()
> class GUI(object):
>     def GUI_Plot(self,widget,event,data=None):
>         self.Graph([TIME,VOLT2])
>
>     def delete_event(self,widget,event,data=None):
>         return False
>
>     def destroy(self,widget,data=None):
>         gtk.main_quit()
>
>
>     def __init__(self):
>         super(GUI,self).__init__()
>
>         self.window = gtk.Window(gtk.WINDOW_TOPLEVEL)
>         self.window.set_title("Matplotlib GTK test")
>         self.window.connect("delete_event",self.delete_event)
>         self.window.connect("destroy",self.destroy)
>
>         self.window.set_border_width (10)
>         self.box1 = gtk.VBox(False,0)
>         self.window.add(self.box1)
>
>         self.button = gtk.Button("GUI Plot")
>         self.button.connect("clicked",self.GUI_Plot,None)
>         self.box1.pack_start(self.button,True,True,0)
>
>
>         self.button.show()
>         self.box1.show()
>         self.window.show()
>
>         self.Graph([TIME,VOLT])
>
>
>     def Graph(self,DATA):
>         try:
>             self.canvas.destroy()
>             self.toolbar.destroy()
>         except:pass
>
>         self.figure = Figure(figsize=(6,3), dpi=100)#{{{
>         self.axis = self.figure.add_subplot(111)
>         self.axis.grid(True)
>         self.axis.set_xlabel('Time (s)')
>
>
>         self.axis.plot(DATA[0],DATA[1],linewidth=2.0)
>
>                  self.canvas = FigureCanvas( self.figure) # a
> gtk.DrawingArea
>             self.canvas.show()
>             self.graphview = self.box1
>
>             self.graphview.pack_start(self.canvas, True, True)
>
>
>
>     def main(self):
>         gtk.main()
>
> if __name__ == "__main__":
>     A = GUI()
>     B = my_thread(A)
>     B.start()
>
>     gtk.threads_enter()
>     A.main()
>     gtk.threads_leave ()
>
>

>>>>> "Nicolas" == Nicolas Bigaouette <nbi...@sy...> writes:
    Nicolas> Clearly, matplotlib's SVG backend doesn't seems to
    Nicolas> support TeX characters. Is it supposed to? 

It does -- the only trick is you need to make sure your svg viewer can
see the fonts.  Currently we use the bakoma cm*.ttf fonts that ship
with matplotlib, so make sure these fonts are in your svg viewer's
font path.  In postscript, we embed the truetype fonts directly into
the PS file which makes for large output files but helps portability
across ps viewers.  We were unable to figure out how to do this with
SVG.  

    Nicolas> Is it possible to hope it will?

One can always hope :-)

JDH

>>>>> "Kevin" == Kevin Horton <kho...@ro...> writes:

    Kevin> I was doing some Googling to look for a way to solve a
    Kevin> problem controlling sharex, and came across an example
    Kevin> using axprops.  This solved my problem, but now I wonder
    Kevin> what other aspects I can control with axprops, or other
    Kevin> similar methods.  Where is axprops documented?  I've looked
    Kevin> via pydoc, and in the pdf documentation, but no dice.

axprops is simply a dictionary holding key/value pairs.  It is not
part of the matplotlib API.  Any function that takes keyword
arguments, such as the Axes constructor, can take a dictionary with
keyword/value pairs using the following syntax

  a = Axes(fig, rect, **d)

where d is a dictionary.  This is part of python, not matplotlib
proper, but because matplotlib makes extensive use of keyword
arguments, it is a handy trick to remember.  When I am creating
several axes with shared properties, I often use it to have a single
customization point

axprops = dict(axisbg='yellow', xlim=(0,1))

for i in range(N):
   fig.add_subplot(N,1,i+1, **axprops)

or something like that.

But I don't think this solves your problem: you can use this to turn
on the sharex feature but not to turn it off once it is already on.

As for your question about where to find the aspects of the Axes that
can be controlled this way, you can do it by consulting the class
documentation at http://matplotlib.sf.net/matplotlib.axes.html and
looking for methods that start with "set_" or by firing up an
interactive shell (see http://matplotlib.sf.net/interactive.html) and
using setp introspection

In [2]: ax = subplot(111)

In [3]: setp(ax)
    adjustable: ['box' | 'datalim']
    alpha: float
    anchor: ['C', 'SW', 'S', 'SE', 'E', 'NE', 'N', 'NW', 'W']
    animated: [True | False]
    aspect: ['auto' | 'equal' | aspect_ratio]
    autoscale_on: True|False
    axis_bgcolor: any matplotlib color - see help(colors)
    axis_off: void
    axis_on: void
    axisbelow: True|False
    clip_box: a matplotlib.transform.Bbox instance
    clip_on: [True | False]
    cursor_props: a (float, color) tuple
    figure: a Figure instance
    frame_on: True|False
    label: any string
    lod: [True | False]
    navigate: True|False
    navigate_mode: unknown
    position: len(4) sequence of floats
    title: str
    transform: a matplotlib.transform transformation instance
    visible: [True | False]
    xlabel: str
    xlim: len(2) sequence of floats
    xscale: ['log' | 'linear' ]
    xticklabels: sequence of strings
    xticks: sequence of floats
    ylabel: str
    ylim: len(2) sequence of floats
    yscale: ['log' | 'linear']
    yticklabels: sequence of strings
    yticks: sequence of floats
    zorder: any number

I was doing some Googling to look for a way to solve a problem  
controlling sharex, and came across an example using axprops.  This  
solved my problem, but now I wonder what other aspects I can control  
with axprops, or other similar methods.  Where is axprops  
documented?  I've looked via pydoc, and in the pdf documentation, but  
no dice.

Thanks,

Kevin Horton
Ottawa, Canada

On Monday 04 September 2006 14:28, Nicolas Bigaouette wrote:
> Hi,
> What is causing this? Did something changed that I need to adapt my code
> or is it a bug?

Check the "aspect" keyword of your image, and try to set it to 'auto'.

> Clearly, matplotlib's SVG backend doesn't seems to support TeX
> characters. Is it supposed to? 
 Yes. 
http://matplotlib.sourceforge.net/matplotlib.texmanager.html

> Is it possible to hope it will? 
Pass

On Monday 04 September 2006 12:15 pm, John Hunter wrote:
> >>>>> "Darren" == Darren Dale <dd...@co...> writes:
>
>     Darren> It is strongly discouraged to try to build matplotlib from
>     Darren> source on windows.  Setting up your windows environment to
>     Darren> build the source is difficult and time consuming. Please
>     Darren> use the windows installers at the download site instead.
>
> I wouldn't go as far to say it is strongly discouraged; merely that it
> is not for the faint of heart.  For enterprising souls who know there
> way around a compiler, I would encourage it, since we would have more
> hands to help with building and testing svn on windows.  As for the
> specific error messages, it looks like you do not have the tk
> development headers installed or for some reason your build is not
> finding them.  Please see the header for win32 in setupext.py and
> references to the READMEs therein if you want to pursue this.

My mistake. Not for the faint of heart is a much better way of putting it.

Hi,

I have 2 questions.

First, after upgrading from 0.86 to 0.87, my graphics aren't the same. 
On 0.87, they are "squeezed" horizontally, which is bad. You can see an 
exemple on those (SVG exported to PNG via Inkscape):
http://nbigaouette.inrs-emt.homelinux.net/linux/matplotlib_0.86.2.png
http://nbigaouette.inrs-emt.homelinux.net/linux/matplotlib_0.87.4.png

The original SVGs are :
http://nbigaouette.inrs-emt.homelinux.net/linux/matplotlib_0.86.2.svg
http://nbigaouette.inrs-emt.homelinux.net/linux/matplotlib_0.87.4.svg

What is causing this? Did something changed that I need to adapt my code 
or is it a bug?


Second, as you can see on the PNG, I can't see well the TeX characters 
(look at the "x" axis, it should be "x (um)" with "u" being mu, the 
micron symbol http://en.wikipedia.org/wiki/Mu_%28letter%29). Those 
graphics are saved directly to SVG, without any GUI. If I display the 
graphic on my screen (without saving to SVG), I get that screen :
http://nbigaouette.inrs-emt.homelinux.net/linux/matplotlib_0.86.2_X.png
If I then save the file to PNG, I get this one :
http://nbigaouette.inrs-emt.homelinux.net/linux/matplotlib_0.86.2_X_save.png
witch is ok. But if I save to SVG, I get this one :
http://nbigaouette.inrs-emt.homelinux.net/linux/matplotlib_0.86.2_X_save.svg
Exported to PNG, the latter is :
http://nbigaouette.inrs-emt.homelinux.net/linux/matplotlib_0.86.2_X_save_svg.png

Clearly, matplotlib's SVG backend doesn't seems to support TeX 
characters. Is it supposed to? Is it possible to hope it will?

Thanks!

Nicolas

>>>>> "David" == David Huard <dav...@gm...> writes:

    David> Hi all, Is there a way to align text vertically so that the
    David> characters sit on a continuous line ?  When I do text(.5,
    David> .5, 'abc', verticalalignment='center') text(.6, .5, 'pqr',
    David> verticalalignment='center') both words are not aligned
    David> since the bars of p and q push the text upwards.  I get
    David> similar problems with 'top' and 'bottom' alignment.

Sadly not.  We need to add an additional alignment flag to support
this.  Currently, we align by the bounding box rather than the text
baseline.

JDH

Hi all,

Is there a way to align text vertically so that the characters sit on a
continuous line ?
When I do
text(.5, .5, 'abc', verticalalignment='center')
text(.6, .5, 'pqr', verticalalignment='center')
both words are not aligned since the bars of p and q push the text upwards.
I get similar problems with 'top' and 'bottom' alignment.


Thanks,

David

Jon Roadley-Battin wrote:
> The problem I have with with the non-GUI thread calling the "plot"
> function.

That's your problem -- you should only call the GUI code (including
matplotlib, which eventually calls GTK) from a single thread. I suggest
looking at the Queue module to pass your data to the GUI thread if it's
not too much. If it's more, you could use threading.Lock() to regulate
access to a giant array or something.

-Andrew

>>>>> "Michael" == Michael Fitzgerald <mp...@be...> writes:


    Michael> I placed that command before the ax.get_xticklabels(),
    Michael> and no dice.

    Michael> Thanks for looking into this, Mike

I took a look at the formatter code and it turns out it *does* know
the list of locations it has to format.  I was wrong about this in my
previous post.  The formatter has a locs attribute you can inspect to
see how many ticks there are.  So you should be able to do something
like

class MyFormatter(ScalarFormatter):
    def __call__(self, x, pos=None):
        N = len(self.locs)
        if pos==0: return ''        # turn off first
        elif pos==(N-1): return ''  # turn off last
        else: return ScalarFormatter(self, x, pos)

and you can do other things similarly, eg to turn off every other tick

if (pos%2)==0: return ''

If you want to get very clever with turning on and off certain ticks,
you can also create a custom locator derived from the Locator class
you are using.

JDH

>>>>> "Samuel" == Samuel GARCIA <sg...@ol...> writes:

    Samuel> Hello, I am writing a little GUI with PyQT4 and
    Samuel> matplotlib.  So it is embeded.  I want that feature for
    Samuel> the user : he can decided after ploting some data to
    Samuel> synchronize 2 graph with the x axis for the zoom.

    Samuel> So this method would be very useful for me. Maybe I can
    Samuel> write it.  Do you have a idea of all the variables
    Samuel> involved in the sharex feature ?

It doesn't look too easy, without some additional methods in the BBox
extension code.  I'll look into it.

JDH

>>>>> "Darren" == Darren Dale <dd...@co...> writes:

    Darren> It is strongly discouraged to try to build matplotlib from
    Darren> source on windows.  Setting up your windows environment to
    Darren> build the source is difficult and time consuming. Please
    Darren> use the windows installers at the download site instead.

I wouldn't go as far to say it is strongly discouraged; merely that it
is not for the faint of heart.  For enterprising souls who know there
way around a compiler, I would encourage it, since we would have more
hands to help with building and testing svn on windows.  As for the
specific error messages, it looks like you do not have the tk
development headers installed or for some reason your build is not
finding them.  Please see the header for win32 in setupext.py and
references to the READMEs therein if you want to pursue this.

JDH

>>>>> ">" ==   <nbi...@sy...> writes:

    >> line 154, in draw_image image64 = base64.b64encode
    >> (imfile.read()) AttributeError: 'module' object has no
    >> attribute 'b64encode'

    >> Is there something I'm missing ?

It looks like the svg module was written assuming the python2.4 version
of the base64 standard library -- the b64encode method does not appear
to be available in python2.3.  I replaced this with base64.encodestring
which appears to work under python2.3 and 2.4.  If you have access to
svn, you can simply update.  Otherwise replace "b64encode" with
"encodestring" in backend_svg.  

JDH

On 4 Sep 2006, at 10:54, John Hunter wrote:

>>>>>> "Kevin" == Kevin Horton <kho...@ro...> writes:
>     Kevin> subplot accepts a tuple in other places, so I would have
>     Kevin> expected the same behaviour here.  Is this a bug that is
>     Kevin> fixed in a newer matplotlib version, in which case I will
>     Kevin> try to upgrade.  Or, is there some other way to do my code
>     Kevin> that will work?
>
> You need to either do
>
>   subplot(10,1,1)
>
> or
>
>   tup = 10,1,1
>   subplot(*tup)
>
> subplot does not accept a tuple: it either accepts and integer, eg
>
>   num = 311
>   subplot(num)
>
> or three args: numrows, numcols, num
>
> python let's you "unpack" tuple with the "*" operator.

That works.  Thank you very much.

Kevin Horton
Ottawa, Canada

Dave,

I had the same problem as you.  Here is a solution...

Try editing your matplotlib rc file and change these font preferences

   font.serif          : New Century Schoolbook, Century Schoolbook L,=20
Utopia, ITC Bookman, Bookman, Bitstream Vera Serif, Nimbus Roman No9 L,=20
Times New Roman, Times, Palatino, Charter, serif
   font.sans-serif     : Lucida Grande, Verdana, Geneva, Lucida,=20
Bitstream Vera Sans, Arial, Helvetica, Avant Garde, sans-serif
   font.cursive        : Apple Chancery, Textile, Zapf Chancery, Sand,=20
cursive
   font.fantasy        : Comic Sans MS, Chicago, Charcoal, Impact,=20
Western, fantasy
   font.monospace      : Andale Mono, Bitstream Vera Sans Mono, Nimbus=20
Mono L, Courier New, Courier, Fixed, Terminal, monospace

by moving the bitstream fonts to the front.

Next, you have to use LaTeX for all the 'text' you include in your plot.=20
  Add rc('text', usetex=3DTrue) at the begining of your Python script or=20
modify your matplotlibrc file.

Everything is working fine after that.

Regards.

Benoit

David Andrews a =E9crit :
> Hiya, i'm having a problem getting a valid postscript file produced on =
OS X.
>=20
> Using the example subplot_demo.py, modified to include the line:
>=20
>     savefig('subplot_demo')
>=20
> in the place of the show() command, and running it using:
>=20
>     python subplot_demo.py -dPS
>=20
> Produces the .ps file (available here for reference:
> http://www.irbdavid.com/misc/subplot_demo.ps ), which can't be opened
> by Preview on os x - says its can't convert it to PDF & cant open the
> file.  Checked the permissions etc on the file and they're fine.
>=20
> Its pretty much a default installation of matplotlib etc, and output
> displays fine on screen, using the wxPython / wxAgg thing (I think
> that's what i mean :D )
>=20
> Any suggestions?
>=20
> Regards,
>=20
> Dave
>=20
> -----------------------------------------------------------------------=
--
> Using Tomcat but need to do more? Need to support web services, securit=
y?
> 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 Geron=
imo
> 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
>=20

--=20
Benoit Donnet
Universite Catholique de Louvain
Facultes des Sciences Appliquees - Departement d'Ingenierie Informatique=20
(INGI)
Place Sainte Barbe, 1
1348 Louvain-la-Neuve
phone: +32 10 47 87 18
home page: http://rp.lip6.fr/~donnet
traceroute@home website: http://trhome.sourceforge.net

>>>>> "David" == David Andrews <irb...@gm...> writes:

    David> Hiya, i'm having a problem getting a valid postscript file
    David> produced on OS X.  Using the example subplot_demo.py,
    David> modified to include the line:

    David>     savefig('subplot_demo')

    David> in the place of the show() command, and running it using:

    David>     python subplot_demo.py -dPS

    David> Produces the .ps file (available here for reference:
    David> http://www.irbdavid.com/misc/subplot_demo.ps ), which can't
    David> be opened by Preview on os x - says its can't convert it to
    David> PDF & cant open the file.  Checked the permissions etc on
    David> the file and they're fine.

    David> Its pretty much a default installation of matplotlib etc,
    David> and output displays fine on screen, using the wxPython /
    David> wxAgg thing (I think that's what i mean :D )

We've noticed this on tiger several times.  Some default system fault
is causing troubles though we haven't identified which one.  The
recommended fix it to put the Vera fonts (which matplotlib ships with)
first in your rc file (http://matplotlib.sf.net/matplotlibrc) and
remove your ~/.matplotlib/ttf.cache before rerunning

Ie, you fonts section of ex should look something like this:


  font.serif          : Bitstream Vera Serif, New Century Schoolbook, Century Schoolbook L, Utopia, ITC Bookman, Bookman, Nimbus Roman No9 L, Times New Roman, Times, Palatino, Charter, serif
  font.sans-serif     : Bitstream Vera Sans, Lucida Grande, Verdana, Geneva, Lucid, Arial, Helvetica, Avant Garde, sans-serif
  font.cursive        : Apple Chancery, Textile, Zapf Chancery, Sand, cursive
  font.fantasy        : Comic Sans MS, Chicago, Charcoal, Impact, Western, fantasy
  font.monospace      : Bitstream Vera Sans Mono, Andale Mono, Nimbus Mono L, Courier New, Courier, Fixed, Terminal, monospace

In future releases of mpl, this will be the default.

JDH

>>>>> "Kevin" == Kevin Horton <kho...@ro...> writes:
    Kevin> subplot accepts a tuple in other places, so I would have
    Kevin> expected the same behaviour here.  Is this a bug that is
    Kevin> fixed in a newer matplotlib version, in which case I will
    Kevin> try to upgrade.  Or, is there some other way to do my code
    Kevin> that will work?

You need to either do

  subplot(10,1,1)

or 

  tup = 10,1,1
  subplot(*tup)

subplot does not accept a tuple: it either accepts and integer, eg

  num = 311
  subplot(num)

or three args: numrows, numcols, num

python let's you "unpack" tuple with the "*" operator.

JDH

>>>>> "Michael" == Michael Fitzgerald <mp...@be...> writes:

    Michael> Hi all,

    Michael> I have a question about zorder and Axis elements.  I read
    Michael> in the thread titled "zorder not working for grid
    Michael> lines??" that since grid lines are Axis elements, they
    Michael> don't respect the zorder.  It's possible to set grid
    Michael> lines below other plot elements with
    Michael> Axis.set_axisbelow().  However, I have a case where I
    Michael> want grid lines to be below my plot elements, but I want
    Michael> the ticks to be above them (test code attached).  As it
    Michael> stands, they can be either both below or both above.

    Michael> Has anyone found a way around this limitation?  Is having
    Michael> zorder apply to Axis elements on the wishlist?

It would be a cleaner solution if the grid and ticks respected the
zorder property, but it's not trivial to do it that way given the
current design.  Instead, as you've noticed, you need to set the
axisbelow property on the axes

  ax.set_axisbelow(True)

but this doesn't solve your problem of needing the ticks above and
grid below....  You might try setting both above, but setting the
alpha on the grid so you can at least see through it.

Yes, fixing this is on the list of things to do, but doing it right
would requires a significant refactoring of the way ticks are
handled.  

JDH

I have run into a problem with matplotlib, GTK and Threading.

The program basically gets data over RS232 and if a certain command is sent
the data is plotted
So far the program is pretty mature and data can be read via RS232 and
written to RS232 and the
GUI is updated (text-widgets) accordingly

The problem I have with with the non-GUI thread calling the "plot" function.

Within the GTK-GUI class I added a method that basically plots a SINE-wave
when a button is pressed it plots another SINE-wave, basically just
proof-of-concept before I got onto the date via RS232

The thread-code to extract data (for plotting) via RS232 works fine but when
the thread calls a GUI method to update the
graph it locks up

CODE-SNIPPITS

class low_level(threading.Thread):
    def __init__(self,Reg,GUI):
        super(low_level, self).__init__()
        self.timer = threading.Timer(3.0,self.__SetTimerState)
        self.Reg = Reg
        self.GUI = GUI
        self.RUN = True
        self.timer_exp = False
        self.LOCAL_LIST = []
        self.WRITELIST = []
...

    def __Capture(self):
        DATA = [[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[]]
        self.__Write(self.Reg.MON_Addr,self.Reg.MON[-1])
        tmp = [x for x in self.Reg.CAPTURE_REQ if x >0 and x not in
locals()["_[1]"]]
        tmp.insert(0,0)

        for i in range(int(self.Reg.CAPTURE_SAMPLES)):
            self.__Write(127)
            for x in tmp:
                t = self.__ReadData(self.Reg.MON_MASK[x][1])
                if t == '::ERROR::':
                    print "Plotting data fault!!"
                    break
                DATA[x].append(t)

        DAT = [DATA[0]]
        for x in self.Reg.CAPTURE_REQ:
            if x <1: DAT.append(None)
            else: DAT.append(DATA[x])

        gtk.threads_enter()#updating GUI section
        try:
            self.GUI.ThreadGraph(DAT)

        finally:
            gtk.threads_leave()
        self.Reg.CAPTURE = False


...
...
...

gtk.gdk.threads_init()
class appgui(object):
    def __init__(self,Reg):
        super(appgui, self).__init__()
        self.Reg = Reg
        self.gladefile=GLADE_FILE
        self.windowname='ESC'
        self.wTree=gtk.glade.XML (self.gladefile,self.windowname)

        dic = {'on_quit_button_clicked' : self.QuitClicked,##callback
dictionary
        'on_connect_button_clicked' : self.Connect,
        'on_save_conf_clicked' : self.Save,
        'on_load_conf_clicked' : self.Load,
        'on_upload_clicked' : self.Upload,#GetRW,
        'on_capture_clicked' : self.Capture,
        'on_startstop_clicked' : self.StartStop,
        'on_window_destroy' : (gtk.main_quit)}

        self.wTree.signal_autoconnect(dic)
        self.f =  self.wTree.get_widget

        self.GraphData(INIT_G)

...
    def GraphData(self,DATA):
        COLOUR = ['b','g','r','c','m','y','k']
        try:
            self.canvas.destroy()
            self.toolbar.destroy()
        except:pass


        self.figure = Figure(figsize=(6,3), dpi=100)#{{{
        self.axis = self.figure.add_subplot(111)
        self.axis.grid(True)
        self.axis.set_xlabel('Time (s)')

        for x in range(1,len(DATA)):
            if DATA[x] == None: continue
            self.axis.plot(DATA[0],DATA[x],COLOUR[x-1],linewidth=2.0)

            self.canvas = FigureCanvas(self.figure) # a gtk.DrawingArea
            self.canvas.show()
            self.graphview = self.wTree.get_widget('vboxgraph')
            self.graphview.pack_start(self.canvas, True, True)
        self.toolbar = NavigationToolbar(self.canvas,self.wTree.get_widget
('ESC'))
        self.graphview.pack_start(self.toolbar, False, False)#}}}


...
if __name__ == '__main__':
    Reg = Register()
    app=appgui(Reg)
    LL = low_level(Reg,app)
    LL.start()

    gtk.threads_enter()
    gtk.main()
    gtk.threads_leave()

    LL.stop()





As it stands with this locks up, if I change the code such that when I press
the widget that starts the "Capture" fn from the thread, BUT that doesn't
update the plot
and if I add another button and in pressing that (NOTE that button's code
would be part of the GUI-class) calls the GraphData function then it works.

It seems even using the gtk.threads_enter and gtk.threads_leave with a
matplotlib embedded into a GTK windows does not work and any graph
modifications MUST
be called directly from the GUI-Class.

Any idea how to make matplotlib work well with GTK & Threads?

I'm a python newbie, using matplotlib 0.87.3.  My code works, as long  
as I limit myself to 9 subplots.  But, I need more than 9.

Relevant parts of the code:

from matplotlib.figure import Figure
import pylab
f = Figure(figsize=(5,4), dpi=100)
plot1 = f.add_subplot((10,1,1), ylabel='Pitch (' + degreeChar + ')')

I get an error like:
Traceback (most recent call last):
   File "Flt_Test_Data_Review.py", line 726, in ?
     create_mpl_plots()
   File "Flt_Test_Data_Review.py", line 246, in create_mpl_plots
     plot1 = f.add_subplot((10,1,1), ylabel='Pitch (' + degreeChar +  
')')
   File "/sw/lib/python2.4/site-packages/matplotlib/figure.py", line  
472, in add_subplot
     a = Subplot(self, *args, **kwargs)
   File "/sw/lib/python2.4/site-packages/matplotlib/axes.py", line  
4316, in __init__
     SubplotBase.__init__(self, fig, *args)
   File "/sw/lib/python2.4/site-packages/matplotlib/axes.py", line  
4208, in __init__
     raise ValueError('Argument to subplot must be a 3 digits long')
ValueError: Argument to subplot must be a 3 digits long

subplot accepts a tuple in other places, so I would have expected the  
same behaviour here.  Is this a bug that is fixed in a newer  
matplotlib version, in which case I will try to upgrade.  Or, is  
there some other way to do my code that will work?

Thanks,

Kevin Horton
Ottawa, Canada

On Monday 04 September 2006 05:20, Paul-Michael Agapow wrote:
> I'm having some problems installing matplotlib (actually PyLab, see
> below). Googling has revealed nothing, perhaps someone might
> recognise these symptoms.

mmh, you didn't check the mailing list before, did you ;)
You didn't precise the version of numpy you were using, I gonna guess that 
it's at least 1.0b1, right ? MPL 0.87.4 is NOT compatible with this version 
of numpy. However, the svn version of matplotlib is, and works quite well.
Please upgrade MPL, or wait a couple of days for the crew to release 0.87.5.

I'm having some problems installing matplotlib (actually PyLab, see  
below). Googling has revealed nothing, perhaps someone might  
recognise these symptoms.

The technical set up: OSX 10.4, MPL 0.87.4, numeric, numarray and  
numpy installed.

While I've successfully installed and used MPL before, this is a new  
MacBook and so it's the first time it's been installed here. (Caveat:  
MPL may have been installed silently in an egg or a package that I  
used to install another package, like a SciPy distribution.) I built  
and installed MPL as usual.  this appears to be fine and `import  
matplotlib` works ok. However the problem comes when I go to `import  
pylab`. Initially it complains that it cannot import pylab because it  
fails on line 74 of  `matplotlib/numerix/__init__.py` where `Matrix =  
matrix`.

Investigation showed that just above that position, numpy was being  
used as the numerical library and the import was thus:

    import numpy.oldnumeric as numpy
    from numpy.oldnumeric import *

However, oldnumeric doesn't define a term "matrix". Setting MPL to  
use Numeric instead of Numpy, this step is passed but I run into the  
next error, also at the "import pylab" stage:

    21:15:23: Debug: ../src/common/object.cpp(224): assert  
"sm_classTable->Get(m_className) == NULL" failed: class already in  
RTTI table - have you used IMPLEMENT_DYNAMIC_CLASS() twice (may be by  
linking some object module(s) twice)?
    ../src/common/object.cpp(224): assert "sm_classTable->Get 
(m_className) == NULL" failed: class already in RTTI table - have you  
used IMPLEMENT_DYNAMIC_CLASS() twice (may be by linking some object  
module(s) twice)?

Initially I thought that this may have been the result of old  
settings hanging around, but I cleaned all vestigaes of MPL & PyLab  
from my system and got a fresh copy of the MPL source,  before trying  
again. Same result.

Any ideas on where I should look next? Apart from a downgrade, I'm at  
a loss.

p

--
Dr Paul-Michael Agapow, VieDigitale / Institute of Animal Health
pm...@vi... / pau...@bb...