matplotlib-users Mailing List for matplotlib

On Fri, Jan 05, 2007 at 04:23:53PM -0600, Glen W. Mabey wrote:
>     p.get_name()
>
> If this last statement returns the first element of
>     matplotlib.rcParams[ 'font.sans-serif' ]
> then I would think that everything is working like it ought to.
> fallback font is being used (fm.defaultFont).

Thanks.  This caught it.  I needed to install a bunch more True Type fonts as I
had too little on my system.


> Contrary to the comments in the default matplotlibrc, it seems that
> font.size does not set the fontsize for axis labels and ticks; you have
> to set [xy]tick.labelsize and axes.labelsize explicitly.

What worked for me was adding fontsize=FONT_SIZE keyword setting
in EVERY command that involved text.

Chris

On Fri, Jan 05, 2007 at 12:21:04PM -0500, ch...@se... wrote:
> In my code I did family="sans-serif" but it didn't seem to have
> any effect.  Any ideas?

I was struggling with this myself this week.  Do:

    import matplotlib.font_manager
    fm = matplotlib.font_manager.FontManager()

If fm.ttfdict doesn't list a bunch of TrueType fonts that you think
should be listed, delete ~/.matplotlib/ttffont.cache , which will cause
it to be rebuilt the next time you import matplotlib.

Then try:

    import matplotlib.text
    t = matplotlib.text.Text( 'alskdjfalksdj' )
    t.get_fontname()
    p = t.get_font_properties()
    p.get_name()

If this last statement returns the first element of 
    matplotlib.rcParams[ 'font.sans-serif' ]
then I would think that everything is working like it ought to.

If you get something like 'vera' at some point, then the absolute-last
fallback font is being used (fm.defaultFont).

Contrary to the comments in the default matplotlibrc, it seems that
font.size does not set the fontsize for axis labels and ticks; you have
to set [xy]tick.labelsize and axes.labelsize explicitly.  But I haven't
had a chance to look into that ... yet.

hope that helps.

Glen

First, I'm new to Matplotlib (and Python!), so apologies in
advance if I'm missing something obvious.

I'm running on XP, Numpy ver. 1.0.1, Matplotlib ver.
0.87.7, Python ver. 2.4.3

I'm trying to create several plots.  I noticed that after
saving each plot my script's memory footprint increases by
about 5MB.  So I quickly run out of system memory.  I
thought maybe I wasn't closing the figure properly.  I
tried the code given in the FAQ and I get the same results.
 I tried calling gc.collect() but it didn't help.  I
noticed that after calling gc.collect(), gc.garbage returns
359 <Tkinter.PhotoImage> objects.  I'm guessing these are
the bad boys who won't go away.  Why are they still there
and why aren't they being "collected"?

Any ideas?

Thanks in advance!

ps.  For what it's worth, here's the code (simplified)
looks like... nothing fancy:

                figure()
                title("Fun")
                (a,b,c) = hist(data, 10)
                c[5].set_facecolor('r')
                savefig(os.path.join("Plots", paramName))
                close()

...like I said, the code from the "memory leak" FAQ also
seems to give me the same problem.

I wish I could help more, but I really have no clue about embedding 
images in SVG -- I haven't even used imshow() at all.

> Perhaps there is a way to increase the dpi of the included
> bitmaps? 

A bitmap has a particular size -- period -- it can be re-scaled by 
interpolating, etc, but it has a given native size in pixels.

I think what may be confusing you is that if you take a given bitmap, 
and increase it's dpi -- all you've done is take the same pixels and 
tell whatever is doing your rendering that each pixel is smaller (so the 
whole image is smaller in length units (inches, etc). For example:

if you have an 800x600 bitmap at 100dpi, you have a 8"x6" picture. If 
you make it 200 dpi, you now have a 4"x3" picture, with smaller dots.

Does that help any?

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/OR&R            (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chr...@no...

In my code I did family="sans-serif"
but it didn't seem to have
any effect.  Any ideas?

Chris

Philippe Fremy
first, install the unicode version of wxPython since matplotlib is compiled
using it.
then add that dll in the variable to be passed to  the data_files argument
of setup, something like this:
matplotlibdata_files.append(('',['E:\\Python25\\Lib\\site-packages\\wx-
2.6-msw-unicode\\wx\\wxmsw26uh_vc.dll']))
then in the setup argument, exclude the dll by having this code
fragment:'dll_excludes' : 'wxmsw26uh_vc.dll',
don't forget to put your data_files argument in setup

HTH
Allan

>
> Message: 9
> Date: Fri, 05 Jan 2007 12:36:06 +0100
> From: Philippe Fremy <ph...@fr...>
> Subject: [Matplotlib-users] matplotlib and wxmsw26uh_vc.dll
> To: mat...@li...
> Message-ID: <459...@fr...>
> Content-Type: text/plain; charset=ISO-8859-1
>
>
>         Hi,
>
>
> This seem to be a recurring problem but I haven't found the answer to my
> problem by googling.
>
> I am packaging an application that uses python 2.5 (from python.org) ,
> ansi version of wxpython
> (version >= 2.6) and numpy 1.01, all this on windows.
>
> When I run my program, it complains that it can't find wxmsw26uh_vc.dll
> but it will run perfectly,
> which I find quite surprising.
>
> I traced the program with dlldpends, and it turns out that it stems from
> the wx backend: _wxagg.pyd
> depends on wxmsw26uh_vc.dll
>
> I did not develop the application, I am only trying to package it and
> trying to understand the
> dependencies. The app runs with various versions of wxPython, but only in
> ansi.
>
> The problem really arises when I try to package the applicaiton with
> py2exe. Py2exe won't let me do
> it. I have tried --dll-excludes and --excludes without success.
>
> Is there a proper fix for this ?
>
>         thanks in advance,
>
>         Philippe Fremy
>
>
>
>
> ------------------------------
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share
> your
> opinions on IT & business topics through brief surveys - and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
>
> ------------------------------
>
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>
> End of Matplotlib-users Digest, Vol 8, Issue 9
> **********************************************
>

Dear Chris,

That site really cleared the basic resolution stuff up for me, and it
was pretty much as I expected. Your remark about how imshow() works with
SVG made me wonder. Might it be that the output of imshow() is set in
pixels so that when I increase the dpi of the total figure, the subplots
decrease in their width and height as measured in inches? (Using the
extent parameter to rescale the bitmap?) This could explain the
'inverse' effect I experienced when increasing dpi. It is not limited to
SVG by the way, EPS and PNG have the same problem.

I got into these problems after upgrading from Ubuntu 6.06 to Ubuntu
6.10, thereby also upgrading from MPL 85.2 (?) to MPL 87.5. In the
previous version the SVG outputted would point to a PNG for each subplot
created with imshow(). Now the bitmaps seem to be included in the SVG
itself. Perhaps there is a way to increase the dpi of the included
bitmaps? (Or, if all else fails, some flag to get the old behavior back?)

Thanks very much,

Marius.

Christopher Barker schreef:
> This should help you understand dpi, font-size, etc.
>
> http://www.scipy.org/Cookbook/Matplotlib/AdjustingImageSize
>
> However, I'm not sure how imshow() and SVG work together -- SVG is
> just that -- "Scalable", it doesn't have a set resolution, and I don't
> know what happens when you embed a raster graphic in it -- can someone
> offer an explanation of what MPL does with imshow and SVG?
>
> -Chris
>
>

Hi all, 
have an happy new year!
I want to make the windrose module more mpl compliant, with setp and others, 
then could you say me what is the best way to do so? 
I have tried to subclasse PolarAxes with the code below, but I've got a bad 
result, like you can see in the picture :-(
An idea?

################
import pylab as P
class WindroseAxes( P.PolarAxes ):
    def __init__( self, *args, **kwargs ):
        P.PolarAxes.__init__( self, *args, **kwargs )
        P.box( False )
        self.set_axisbelow( False )
        self.set_thetagrids( angles=arange(0,360,45), 
labels=['E','N-E','N','N-W','W','S-W','S','S-E'], frac=1.2, weight='bold' )
###############

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

	Hi,


This seem to be a recurring problem but I haven't found the answer to my problem by googling.

I am packaging an application that uses python 2.5 (from python.org) , ansi version of wxpython
(version >= 2.6) and numpy 1.01, all this on windows.

When I run my program, it complains that it can't find wxmsw26uh_vc.dll but it will run perfectly,
which I find quite surprising.

I traced the program with dlldpends, and it turns out that it stems from the wx backend: _wxagg.pyd
depends on wxmsw26uh_vc.dll

I did not develop the application, I am only trying to package it and trying to understand the
dependencies. The app runs with various versions of wxPython, but only in ansi.

The problem really arises when I try to package the applicaiton with py2exe. Py2exe won't let me do
it. I have tried --dll-excludes and --excludes without success.

Is there a proper fix for this ?

	thanks in advance,

	Philippe Fremy

Hoi,

For some plot I'd like to display a dotted line (style = 'k.-') with an error 
bar at only every 30th point or so. The error values in this case are 
scalars.
Of course, I could produce a slice like mydata[::jumper] and plot these above 
the first plot, but this seems a bit akward. Anyone with a better solution 
out there?

TIA
Christian

Tommy Grav <tg...@ma...> writes:

> You have to use set_ylim((0,2)) to set the y limits.

Is there a function/method to calculate apropriate ?lim values from a
given Dataset?

Regards
Berthold
-- 
                            __   Address:
                         G /  \ L Germanischer Lloyd
phone: +49-40-36149-7374 -+----+- Vorsetzen 35       P.O.Box 111606
fax  : +49-40-36149-7320   \__/   D-20459 Hamburg    D-20416 Hamburg