matplotlib-users Mailing List for matplotlib (Page 4)

On Friday 23 June 2006 8:12 pm, Johan Fredrik =D8hman wrote:
> Hi,
>
> I made a plot where the values on the X-axis are dates, "JUN, JUL,
> AUG.. etc ". To make them fit better, I have rotated the text using
> setp(labels, rotation=3D45, size=3D8), the problem is that the text is
> "clipped" in bottom.  How can I expand the boundries?  Thanks !!
>

Instead of this...
> ax =3D subplot(111)

Try this...
fig =3D figure(num=3D1, figsize=3D(2,2))
ax =3D fig.add_axes([left, bottom, width, height])=20

Start with left=3D0.1, bottom=3D0.1, width=3D0.8, height=3D0.8 and see how=
=20
that looks. Adjust the values of bottom and/or height until your=20
labels are no longer clipped.=20

Hi,

I made a plot where the values on the X-axis are dates, "JUN, JUL, AUG.. et=
c ".
To make them fit better, I have rotated the text using setp(labels,
rotation=3D45, size=3D8), the problem is that the text is "clipped" in bott=
om.  How
can I expand the boundries?  Thanks !!

figure(num=3D1, figsize=3D(2,2))
#xlabel('Tid')
ylabel('Temp', size=3D8)
title('Frysertemperatur', size =3D 10)
p =3D plot(dates, s)
ax =3D subplot(111)
ax.set_clip_on(False)
ax.xaxis.set_major_locator(days)
ax.xaxis.set_major_formatter(daysFmt)
#ax.xaxis.set_minor_locator(hours)
labels =3D ax.get_xticklabels()
setp(labels, rotation=3D45, size=3D8)
labels =3D ax.get_yticklabels()
setp(labels, size=3D8)
grid(True)
savefig('simple_plot')
show()

Sincerely
JF=D8

John
> Well, we could keep it simple and just give a hook to custom figures.
> If a user wants a custom subplot
>
> class MyFigure(Figure):
>    def add_my_subplot(self, *args, **kwargs):
>         self.axes.append(MySubplot(*args, **kwargs))
>

> Is there any downside to this approach?  It seems like if you are
> going to the trouble to create a custom figure, you might was well
> handle your own custom subplots there rather than in the Figure class
> proper.

Try that:
fig = figure(FigureClass=MyFigure)
fsplist = [Subplot(fig,211), Subplot(fig,212), Subplot(fig,211)]
for fsp in fsplist: 
	fig.add_my_subplot(fsp)

You should end up with three subplots, with fsp[-1] and fsp[0] at the same 
location 211, when in fact you really need only two subplots (assume that the 
last one is an updated version of the first one, which is not obvious in this 
example....). In short, the `add_my_subplot` does not reproduce the behaviour 
of `add_subplot`.

FYI, when I started working on that yesterday evening, I wrote something quite 
similar to what you're suggesting. Then, I needed to go through all my 
subplots, and instead of checking  the `figure.axes` list, I checked 
`fig._axstack_elements`. Which was empty, of course. So I added a 
`self._axstack.push(MySubplot`) to `add_my_subplot`, then noted that I had to 
put back the tests on whether the dubplot was already present or not 
(figure._seen), and before soon, I was just copying the basic 'add_subplot()' 
around that `self.axes.append(MySubplot)`. As I have several custom figures 
with custom subplots, I realized it'd be cleaner to just make a 
`add_generic_subplot` instead of copying/pasting most of the code.

I'm realizing now that had I worked on `figure.axes` straight away, I would 
never have had the proper behaviour, so I've been lucky.

Hi,

What's a good way to create a 2d gaussian kernel with pylab?

I want a 9x9 matrix with 1.0 in the center, and 0.0 in the corners,
and a gaussian distribution from the center out.

Thanks,
-Jim

Hi,

What's a good way to create a 2d gaussian kernel with pylab?

I want a 9x9 matrix with 1.0 in the center, and 0.0 in the corners,
and a nice gaussian distribution from the center out.

Thanks!
-Jim

>>>>> "PGM" == PGM  <pgm...@gm...> writes:

    PGM> Now, till we (you ;)) are it, what about updating
    PGM> Figure.add_subplot() to call user-defined subplots ? I wrote
    PGM> something along the lines below for my own needs, but that
    PGM> might be a useful addition...

Well, we could keep it simple and just give a hook to custom figures.
If a user wants a custom subplot

class MyFigure(Figure):
   def add_my_subplot(self, *args, **kwargs):        
        self.axes.append(MySubplot(*args, **kwargs))

fig = figure(FigureClass=MyFigure)
fig.add_my_subplot(111)


Is there any downside to this approach?  It seems like if you are
going to the trouble to create a custom figure, you might was well
handle your own custom subplots there rather than in the Figure class
proper.

JDH

John

> I modified pylab (and every backend, damn there are a lot) to support
> this feature.  

Wow, impressive ! Thx a lot

> Now you can pass a FigureClass kwarg to the pylab 
> figure function.  With minimal extra work, we could support defaults
> so you don't have to explicitly pass it.  But before I go ahead with
> this, take a look and see if this is more like what you had in mind.

Yep indeed, that's pretty much what I was considering. And parsing the 
MyFigure arguments through keywords seems the easiest, I'll keep that in mind 
(I used optional arguments *args, as they're not used in Figure, but that's 
trickier to keep track of).

Now, till we (you ;)) are it, what about updating Figure.add_subplot() to call 
user-defined subplots ? I wrote something along the lines below for my own 
needs, but that might be a useful addition...

Initial add_subplot
add_subplot(self, *args, **kwargs):
	...
        if isinstance(args[0], Subplot) or isinstance(args, PolarSubplot):
            a = args[0]
            a.set_figure(self)
        else:
            ispolar = popd(kwargs, 'polar', False)
            if ispolar:
                a = PolarSubplot(self, *args, **kwargs)
            else:
                a = Subplot(self, *args, **kwargs)
	...

Modified add_subplot
add_generic_subplot(figure, subplotclass, *args, **kwargs):
	...
        if isinstance(args[0], Subplot):
            a = args[0]
            a.set_figure(self)
        else:
            a = subplotclass(self, *args, **kwargs)
	...

That way, if you need a PolarSubplot, you can still get it. Well, we'd have to 
make PolarSubplot a subclass of Subplot instead of Subplotbase, but that 
should be easy to implmnt.



> from pylab import figure, show, nx
> from matplotlib.figure import Figure
>
> class MyFigure(Figure):
>     def __init__(self, *args, **kwargs):
>         """
>         custom kwarg figtitle is a figure title
>         """
>         figtitle = kwargs.pop('figtitle', 'hi mom')
>         Figure.__init__(self, *args, **kwargs)
>         self.text(0.5, 0.95, figtitle, ha='center')
>
> fig = figure(FigureClass=MyFigure, figtitle='my title')
> ax = fig.add_subplot(111)
> ax.plot([1,2,3])
> fig.savefig('test.png')
> fig.savefig('test.ps')
> show()
>
>
>
>
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Hi John

Thanks for that. I thought I recalled seeing the example you mentioned, 
but obviously it was a different one.

I tried the code you gave me and it worked perfectly. Thanks very much!

Cheers
JP

John Hunter wrote:
> The approach I suggested doesn't affect the x and y axes, only the
> coord in the toolbar.
>   
...
> Try
>
>   def myfmt(x,y): return 'myfmt (%1.2f, %1.2f)'%(x,y)  
>   ax = subplot(111)
>   ax.format_coord = myfmt
>   ax.plot([1,2,3])
>
> JDH
>
>   
>

-- 
John Pye
Department of Mechanical and Manufacturing Engineering
University of New South Wales, Sydney, Australia
http://pye.dyndns.org/

>>>>> "John" == John Pye <joh...@st...> writes:

    John> Hi all I have an application that uses Matplotlib to display
    John> incidence matrices for systems of equations. I have linked
    John> up the figure window with 'event handling' so that mouseover
    John> on the plot shows the variables/relations corresponding to
    John> the pointed-to coordinates, but at present this is only
    John> output to the console, as per the example in the User's
    John> Guide.

    John> I'd like to switch it over so that the variable and relation
    John> name replaces the 'x=' and 'y=' text in the bottom-left of
    John> the Figure window.

    John> Is there anything in the event handling mechanism that would
    John> facilitate me doing this? Is there a way I can do this
    John> without having to patch matplotlib? It seems like this would
    John> be a generally useful addition to the event handling stuff
    John> for figure windows.

The Axes class has two attributes, fmt_xdata and fmt_ydata, which if
defined are callables that take the x and y coords as floats and
return the format string used to format that bit of text that goes
into the toolbar.  The default is to use the xaxis and yaxis
Formatters.  See examples/date_demo1.py for some example code.

JDH

>>>>> "Pierre" == Pierre GM <pgm...@ma...> writes:

    Pierre> John, Thx for your answer. IMHO, the problem lies really
    Pierre> with Figure, not Canvas.  I'd need the Canvas, whatever
    Pierre> backend defines it, recognizes that the figure is not a
    Pierre> classical 'Figure' , but a subclass of it, with its own
    Pierre> special properties/methods.  I gonna think aloud for a
    Pierre> minute

I modified pylab (and every backend, damn there are a lot) to support
this feature.  Now you can pass a FigureClass kwarg to the pylab
figure function.  With minimal extra work, we could support defaults
so you don't have to explicitly pass it.  But before I go ahead with
this, take a look and see if this is more like what you had in mind.  

from pylab import figure, show, nx
from matplotlib.figure import Figure

class MyFigure(Figure):
    def __init__(self, *args, **kwargs):
        """
        custom kwarg figtitle is a figure title
        """
        figtitle = kwargs.pop('figtitle', 'hi mom')
        Figure.__init__(self, *args, **kwargs)
        self.text(0.5, 0.95, figtitle, ha='center')

fig = figure(FigureClass=MyFigure, figtitle='my title')
ax = fig.add_subplot(111)
ax.plot([1,2,3])
fig.savefig('test.png')
fig.savefig('test.ps')
show()
            

Hi all

I have an application that uses Matplotlib to display incidence matrices 
for systems of equations. I have linked up the figure window with 'event 
handling' so that mouseover on the plot shows the variables/relations 
corresponding to the pointed-to coordinates, but at present this is only 
output to the console, as per the example in the User's Guide.

I'd like to switch it over so that the variable and relation name 
replaces the 'x=' and 'y=' text in the bottom-left of the Figure window.

Is there anything in the event handling mechanism that would facilitate 
me doing this? Is there a way I can do this without having to patch 
matplotlib? It seems like this would be a generally useful addition to 
the event handling stuff for figure windows.

Cheers
JP

-- 
John Pye
Department of Mechanical and Manufacturing Engineering
University of New South Wales, Sydney, Australia
http://pye.dyndns.org/

On Tuesday 20 June 2006 11:37, John Hunter wrote:
> >>>>> "Pierre" == Pierre GM <pgm...@ma...> writes:
>
>     Pierre> Folks,
>     Pierre> I have a tailor-made subclass of Figure (say, MyFigure) that I'd
>     Pierre> like to use interactively. Ideally, I'd like to have
>     Pierre> pylab.figure() (or an equivalent) create a MyFigure instance,
>     Pierre>     instead of Figure. 
>     Pierre>
>     Pierre> I wrote a myfigure() function based on pylab.figure(), 
>     Pierre> substituting the 
>     Pierre> -----
>     Pierre> figManager = new_figure_manager(num, **figargs)
>     Pierre> -----
>     Pierre> by
>     Pierre> -----
>     Pierre> thisFig = MyFigure(*args, **kwargs)
>     Pierre> canvas = FigureCanvasGTKAgg(thisFig)
>     Pierre> figManager = FigureManagerGTKAgg(canvas, num)
>     Pierre> ----
>     Pierre>
>     Pierre> That seems to give me what I want, as long as I use the
>     Pierre> GTKAgg backend. What should I do to have the same result
>     Pierre> with another backend, without having to rewrite a
>     Pierre> new_figure_manager each time ? Anyway, is it even the way
>     Pierre> to go ?
>
> Note that pylab basically has the same problem, as each backend
> defines their own new_figure_manager function.  The problem is not the
> Figure, but the FigureCanvas, which is backend dependent.  Take a look
> at the "switch_backend" function in pylab
>
> def switch_backend(newbackend):
>     """
>     Calling this command will close all open windows.
>     """
>     close('all')
>     global new_figure_manager, draw_if_interactive, show
>     matplotlib.use(newbackend)
>     reload(backends)
>     from backends import new_figure_manager, draw_if_interactive, show
>
> You could emulate this approach ....

John, 
Thx for your answer. IMHO, the problem lies really with Figure, not Canvas. 
I'd need the Canvas, whatever backend defines it, recognizes that the figure 
is not a classical 'Figure' , but a subclass of it, with its own special 
properties/methods.
I gonna think aloud for a minute

One way would be to add a 'subclass' option to every new_figure_manager,
so that the 
`thisFig = Figure(**figargs)`
would be replaced by
`thisFig = subclass(**figargs)` 
(with subclass set to Figure by default). A bit overkill, eh ?

In fact, as long as my special properties/methods can be accessed Figure 
instance, I should be OK. What would be the better way to have the 
properties/methods of a child recognized by its parent (revered 
inheritance ?)

can someone tell me why the lines on the plot are drawn with a linewidth 
of 1 instead of 5?


     def init_plot_data(self):
         self.axes = self.fig.add_subplot(111)
         self.xrange = numpy.arange(80)
         self.yrange = range(80)
         random.shuffle(self.yrange)
         yrange = numpy.array(self.yrange)
         self.axes.plot(self.xrange,  self.yrange, lw=5)

thanks,

bryan

>>>>> "Pierre" == Pierre GM <pgm...@ma...> writes:

    Pierre> That seems to give me what I want, as long as I use the
    Pierre> GTKAgg backend. What should I do to have the same result
    Pierre> with another backend, without having to rewrite a
    Pierre> new_figure_manager each time ? Anyway, is it even the way
    Pierre> to go ?

Note that pylab basically has the same problem, as each backend
defines their own new_figure_manager function.  The problem is not the
Figure, but the FigureCanvas, which is backend dependent.  Take a look
at the "switch_backend" function in pylab

def switch_backend(newbackend):
    """
    Swtich the default backend to newbackend.  This feature is
    EXPERIMENTAL, and is only expected to work switching to an image
    backend.  Eg, if you have a bunch of PS scripts that you want to
    run from an interactive ipython session, yuo may want to switch to
    the PS backend before running them to avoid having a bunch of GUI
    windows popup.  If you try to interactively switch from one GUI
    backend to another, you will explode.

    Calling this command will close all open windows.
    """
    close('all')
    global new_figure_manager, draw_if_interactive, show
    matplotlib.use(newbackend)
    reload(backends)
    from backends import new_figure_manager, draw_if_interactive, show

You could emulate this approach ....

JDH

Folks,
I have a tailor-made subclass of Figure (say, MyFigure) that I'd like to use 
interactively. Ideally, I'd like to have pylab.figure() (or an equivalent) 
create a MyFigure instance, instead of Figure.

I wrote a myfigure() function based on pylab.figure(), substituting the
-----
figManager = new_figure_manager(num, **figargs) 
-----
by
-----
thisFig = MyFigure(*args, **kwargs)
canvas = FigureCanvasGTKAgg(thisFig)
figManager = FigureManagerGTKAgg(canvas, num)
----

That seems to give me what I want, as long as I use the GTKAgg backend. What 
should I do to have the same result with another backend, without having to 
rewrite a new_figure_manager each time ? Anyway, is it even the way to go ?

Thanks for any ideas
P.

Ryan Krauss wrote:
> Sorry, I didn't scroll down low enough in the message to see the png
> you already attached.
> 
> Ryan
> 

The latest update (Darren's mail) makes the plots (exponents etc.) look
just cool. Thanks!

cheers,
steve

-- 
Random number generation is the art of producing pure gibberish as
quickly as possible.

I have updated the apt repository I maintain for Ubuntu's Dapper, which
now includes:

numpy
matplotlib
scipy

Each package is from a recent SVN checkout and should thus be regarded
as "bleeding edge". The repository has a new URL:
http://debs.astraw.com/dapper/ I intend to keep this repository online
for an extended duration. If you want to put this repository in your
sources list, you need to add the following lines to /etc/apt/sources.list::
 deb http://debs.astraw.com/ dapper/
 deb-src http://debs.astraw.com/ dapper/

I have not yet investigated the use of ATLAS in building or using the
numpy binaries, and if performance is critical for you, please evaluate
speed before using it. I intend to visit this issue, but I cannot say when.

The Debian source packages were generated using stdeb, [
http://stdeb.python-hosting.com/ ] a Python to Debian source package
conversion utility I wrote. stdeb does not build packages that follow
the Debian Python Policy, so the packages here may be slighly unusual
compared to Python packages in the official Debian or Ubuntu
repositiories. For example, example scripts do not get installed, and no
documentation is installed. Future releases of stdeb may resolve these
issues.

As always, feedback is very appreciated.

Cheers!
Andrew

Hi John

Here's a test script that gives the wrongly exported SVG image. I'm 
viewing with EOG. When I edit to add "style="opacity:0.15"" it works 
fine, so there's no prob with EOG.

from pylab import *
import matplotlib.numerix.ma as ma
from matplotlib.colors import ListedColormap

A = flipud(array([[1,2,3,4,5],[1,3,4,5,6],[1,1,4,5,7]]))

M = where(A==1,1,0)

B = ma.array(A,mask=M)

C = ListedColormap(['r','#FF00FF','y','g','b','#FFFF00'])
x = [0,5,10,15,20]
y = [0,5,10]
X,Y = meshgrid(x,y)
X = ma.array(X,mask=B.mask())
Y = ma.array(Y,mask=B.mask())

figure()
pcolor(X,Y,B,alpha=0.15,shading='flat',cmap=C)
hold(True)
contour(x,y,B)
xlabel('something')
ylabel('else')

Cheers
JP

John Hunter wrote:
>>>>>> "John" == John Pye <joh...@st...> writes:
>>>>>>             
>
>     John> issues: 1. When exporting the following image, which is a
>     John> pcolor plot with alpha=0.15 and contours on top, I get no
>     John> alpha channel in the resulting output, regard of whether I
>     John> export as EPS or SVG. Is exporting with alpha channel not
>     John> supported in those formats?  2. In the titles, which are
>
>
> For some reason I did not get the sample image, but could you provide
> some example code that replicates the alpha problem on the SVG
> backend?  As others have noted, PS doesn't have an alpha channel so
> there is nothing we can do there, but SVG *should* work.  Are you sure
> it's not your SVG viewer?
>
> JDH
>
>
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>   

-- 
John Pye
Department of Mechanical and Manufacturing Engineering
University of New South Wales, Sydney, Australia
http://pye.dyndns.org/

thanks for the suggestion- that works as expected:

 fill(fx, fy, label = '_nolegend_')

a

John Hunter <jdh...@ac...> wrote: >>>>> "aonghus" == aonghus   writes:

    aonghus> Hi, Does anyone know how to use the fill command without
    aonghus> a label appearing in the legend box?

    aonghus> Any suggestions?

>From the legend docstring

  if label is set to '_nolegend_', the item will not be shown in
  legend.


JDH


 		
---------------------------------
Ring'em or ping'em. Make  PC-to-phone calls as low as 1¢/min with Yahoo! Messenger with Voice.

>>>>> "aonghus" == aonghus  <the...@ya...> writes:

    aonghus> Hi, Does anyone know how to use the fill command without
    aonghus> a label appearing in the legend box?

    aonghus> Any suggestions?

>From the legend docstring

  if label is set to '_nolegend_', the item will not be shown in
  legend.


JDH

Hi,

Does anyone know how to use the fill command without a label appearing in the legend box?

I have something like this:

 plot(x, y, label = 'label1')
 fill(fx, fy)
 legend(loc = 'best')

I thought this might work:

 fill(fx, fy, label = None)

but I get this error instead:

 raise TypeError('text must be a string type')
TypeError: text must be a string type


Any suggestions?

thanks,

a


 			
---------------------------------
Sneak preview the  all-new Yahoo.com. It's not radically different. Just radically better. 

>>>>> "John" == John Pye <joh...@st...> writes:

    John> issues: 1. When exporting the following image, which is a
    John> pcolor plot with alpha=0.15 and contours on top, I get no
    John> alpha channel in the resulting output, regard of whether I
    John> export as EPS or SVG. Is exporting with alpha channel not
    John> supported in those formats?  2. In the titles, which are


For some reason I did not get the sample image, but could you provide
some example code that replicates the alpha problem on the SVG
backend?  As others have noted, PS doesn't have an alpha channel so
there is nothing we can do there, but SVG *should* work.  Are you sure
it's not your SVG viewer?

JDH

>>>>> "Ryan" == Ryan Krauss <rya...@gm...> writes:

    Ryan> Thanks John.  The latest svn works beautifully for me now.
    Ryan> No more svn updates for me until my thesis is submitted.

Uh huh, just like the man who says no more drinks for me until my
thesis is submitted.  You know you can't resist :-)

JDH

Thanks John.  The latest svn works beautifully for me now.

No more svn updates for me until my thesis is submitted.

Ryan

On 6/18/06, John Hunter <jdh...@ac...> wrote:
> >>>>> "Ryan" == Ryan Krauss <rya...@gm...> writes:
>
>     Ryan> I just did a fresh svn checkout and my legends are no longer
>     Ryan> in the front.  The following lines produce the attached
>     Ryan> plot:
>
>
> Hey Ryan,
>
> Not sure how or why this bug crept in, but apparently legend was
> missing a zorder.  I just added a default zorder value, which fixes
> this problem.  Thanks for the report.
>
> JDH
>