matplotlib

Chris Barker wrote:
> Hi all,
> 
> I'm having an odd issue with the wxAgg back-end:

Update:

If I remove:

matplotlib/backends/_wxagg.pyd

The problem goes away.

It looks like that pyd is getting loaded even though I'm running wxPython2.8

However, now I get a non-valid png when I do savefig....


-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

Chris.Barker@...

More complete:

I tried all permunations of backends. Now I stick to PS, because I use
matplotlib from commandline with scripts. The environment is debian/etch
with a current version of matplotlib (self compiled).

Try this script
********************************************
from pylab import *

t = arange(0.0, 2.0, 0.01)
s = sin(2*pi*t)
plot(t, s)
grid(True)

# matlab handle graphics style
xticklines = getp(gca(), 'xticklines')
yticklines = getp(gca(), 'yticklines')
xgridlines = getp(gca(), 'xgridlines')
ygridlines = getp(gca(), 'ygridlines')
xticklabels = getp(gca(), 'xticklabels')
yticklabels = getp(gca(), 'yticklabels')

setp(xticklines, 'linewidth', 3)
setp(yticklines, 'linewidth', 3)
setp(xgridlines, 'linestyle', '-')
setp(ygridlines, 'linestyle', '-')
setp(yticklabels, 'color', 'r', fontsize='medium')
setp(xticklabels, 'color', 'r', fontsize='medium', family='sans-serif')

savefig('axprops_demo')
******************************************

On my system the line
 setp(xticklabels, 'color', 'r', fontsize='medium', family='sans-serif')
doesn't change the font family.

Best regards:
 Gerolf
-- 
Dipl. Phys. Gerolf Ziegenhain (gerolf@...)
Private: Klopstockstrasse 21 - 65187 Wiesbaden - Germany
Office: Room 46-332 - Erwin-Schrödinger-Str.46 - TU Kaiserslautern - Germany
Web: gerolf.ziegenhain.com
Tel.: +49-611-18840590
Fax: +49-611-18840599

On Thursday 27 March 2008 01:27:28 pm Gerolf Ziegenhain wrote:
> More complete:
>
> I tried all permunations of backends. Now I stick to PS, because I use
> matplotlib from commandline with scripts. The environment is debian/etch
> with a current version of matplotlib (self compiled).
>
> Try this script
> ********************************************
> from pylab import *
>
> t = arange(0.0, 2.0, 0.01)
> s = sin(2*pi*t)
> plot(t, s)
> grid(True)
>
> # matlab handle graphics style
> xticklines = getp(gca(), 'xticklines')
> yticklines = getp(gca(), 'yticklines')
> xgridlines = getp(gca(), 'xgridlines')
> ygridlines = getp(gca(), 'ygridlines')
> xticklabels = getp(gca(), 'xticklabels')
> yticklabels = getp(gca(), 'yticklabels')
>
> setp(xticklines, 'linewidth', 3)
> setp(yticklines, 'linewidth', 3)
> setp(xgridlines, 'linestyle', '-')
> setp(ygridlines, 'linestyle', '-')
> setp(yticklabels, 'color', 'r', fontsize='medium')
> setp(xticklabels, 'color', 'r', fontsize='medium', family='sans-serif')
>
> savefig('axprops_demo')
> ******************************************
>
> On my system the line
>  setp(xticklabels, 'color', 'r', fontsize='medium', family='sans-serif')
> doesn't change the font family.

Do you use "usetex"? If you do, the mathtext is rendered using latex, and 
latex renders mathtext in serif fonts.

Darren

Hi,

Thanks for all the replies! Of course it is not directly a bug. But
awkward. This is how I was finally able to change alle fonts and the
whole layout of the plot:

fig_width_pt = 246.0  # Get this from LaTeX using \showthe\columnwidth
inches_per_pt = 1.0/72.27               # Convert pt to inch
golden_mean = (sqrt(5)-1.0)/2.0         # Aesthetic ratio
fig_width = fig_width_pt*inches_per_pt  # width in inches
fig_height = fig_width*golden_mean      # height in inches
fig_size =  [fig_width,fig_height]
params = {'backend': 'ps','axes.labelsize': 10,'text.fontsize': 10,'legend.fontsize': 10,'text.usetex': True,'text.latex.preamble':["\usepackage{bm}","\usepackage{textcomp}"], 'figure.figsize': fig_size, 'font.family' : 'sans-serif', 'font.style': 'normal', 'font.weight':'normal',}
rcParams.update(params)

.......

ffont = {'family':'sans-serif','fontsize':10,'weight':'bold'}
ax1=gca()
ax1.set_xticklabels(ax1.get_xticks(),ffont)
ax1.set_yticklabels(ax1.get_yticks(),ffont)
subplots_adjust(top=1-.13)
xticks(**ffont)
yticks(**ffont)

Best regards:
   Gerolf

On Thu 27.Mar.08 14:54, Darren Dale wrote:
> On Thursday 27 March 2008 01:27:28 pm Gerolf Ziegenhain wrote:
> > More complete:
> >
> > I tried all permunations of backends. Now I stick to PS, because I use
> > matplotlib from commandline with scripts. The environment is debian/etch
> > with a current version of matplotlib (self compiled).
> >
> > Try this script
> > ********************************************
> > from pylab import *
> >
> > t = arange(0.0, 2.0, 0.01)
> > s = sin(2*pi*t)
> > plot(t, s)
> > grid(True)
> >
> > # matlab handle graphics style
> > xticklines = getp(gca(), 'xticklines')
> > yticklines = getp(gca(), 'yticklines')
> > xgridlines = getp(gca(), 'xgridlines')
> > ygridlines = getp(gca(), 'ygridlines')
> > xticklabels = getp(gca(), 'xticklabels')
> > yticklabels = getp(gca(), 'yticklabels')
> >
> > setp(xticklines, 'linewidth', 3)
> > setp(yticklines, 'linewidth', 3)
> > setp(xgridlines, 'linestyle', '-')
> > setp(ygridlines, 'linestyle', '-')
> > setp(yticklabels, 'color', 'r', fontsize='medium')
> > setp(xticklabels, 'color', 'r', fontsize='medium', family='sans-serif')
> >
> > savefig('axprops_demo')
> > ******************************************
> >
> > On my system the line
> >  setp(xticklabels, 'color', 'r', fontsize='medium', family='sans-serif')
> > doesn't change the font family.
> 
> Do you use "usetex"? If you do, the mathtext is rendered using latex, and 
> latex renders mathtext in serif fonts.
> 
> Darren
> 
> -------------------------------------------------------------------------
> Check out the new SourceForge.net Marketplace.
> It's the best place to buy or sell services for
> just about anything Open Source.
> http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
> _______________________________________________
> Matplotlib-users mailing list
> Matplotlib-users@...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

-- 
Dipl. Phys. Gerolf Ziegenhain 
Web:     http://gerolf.ziegenhain.com
Email:   mail.gerolf@...
Fax:     +49 611 18840599
Phone:   +49 611 18840590
Office:  Erwin-Schrödinger-Str. 46 / Room 46.332 / 67663 Kaiserslautern / Germany
Private: Klopstockstr. 21 / 65187 Wiesbaden / Germany

Hello Chris,
Hello list,

On Thursday 27 March 2008 18:26, Chris Withers wrote:
> Matthias Michler wrote:
> > I'm not sure that I understand you correctly. The code I refering is the
> > one which I attached some mails ago. The following works for me:
>
> Ah, okay, to get the problem I was having, change your script as follows:
> > -------------------------------------------------------------------------
> >-------- from pylab import *
> >
>  > from datetime import datetime
> >
> > from time import sleep
> >
> > ion()                     # interactive mode 'on'
> > figure()
> > ax = subplot(111, autoscale_on=True)
> >
> > x, y = [datetime.now()], [0]
> > line = plot(x, y, label="my_data")[0]
> >                                     # get the line-object as the first
> > element # of the tuple returned by plot legend()
> > for i in arange(30):
> >     x.append(datetime.now())           # append new values
> >     y.append(i**2)
> >     line.set_data(x,y)    # reset data
> >     ax.relim()            # reset axes limits
> >     ax.autoscale_view()   # rescale axes
> >     draw()                # redraw current figure
> >     sleep(0.3)            # wait 0.3 seconds
> >
> > ioff()
> > show()
>
> So, basically make the x axis time instead of numbers.
> I think the problem is actually that the daets are quite long in their
> format. If they were rotated through 90 degress it'd likely be fine.
> How would I do this?

I'm not sure it is the easiest way, but it works for me:

for label in ax.xaxis.get_majorticklabels():
    label.set_rotation(+90)

> Also, how would I get this kind of updating with bar charts or errorbars?

In the case of bar charts and errorbars it is quite difficult to reset the 
data.  
e.g. errobar (from the docstring) 
Return value is a length 3 tuple.  The first element is the
    Line2D instance for the y symbol lines.  The second element is
    a list of error bar cap lines, the third element is a list of
    line collections for the horizontal and vertical error ranges
I think it is quite expensive to reset all x/ydata of the lines by yourself 
and I have no idea how to reset data of line collections.

I would replot the errorbars, but maybe somebody else knows a good way to 
reset the data of errorbars.

regards 
Matthias

Matthias Michler wrote:
> I'm not sure it is the easiest way, but it works for me:
> 
> for label in ax.xaxis.get_majorticklabels():
>     label.set_rotation(+90)

Yes, that's what I was using, just wondered if there was a better way...

>> Also, how would I get this kind of updating with bar charts or errorbars?
> 
> In the case of bar charts and errorbars it is quite difficult to reset the 
> data.  

Oh, I also meant to ask about scatter, can the data be easilly reset there?

For bar charts and errorbar plots, I agree ;-) How would I just blank 
the figure and replot?
(I have just been calling errorbar lots, but I'm guessing that if I add 
a legend, I'll have one entry for each time I called errorbar :-S)

cheers,

Chris

-- 
Simplistix - Content Management, Zope & Python Consulting
            - http://www.simplistix.co.uk

On Friday 28 March 2008 13:57, Chris Withers wrote:
> Matthias Michler wrote:
> > I'm not sure it is the easiest way, but it works for me:
> >
> > for label in ax.xaxis.get_majorticklabels():
> >     label.set_rotation(+90)
>
> Yes, that's what I was using, just wondered if there was a better way...

At least I don't know a better way, but I'm not an expert.

> >> Also, how would I get this kind of updating with bar charts or
> >> errorbars?
> >
> > In the case of bar charts and errorbars it is quite difficult to reset
> > the data.
>
> Oh, I also meant to ask about scatter, can the data be easilly reset there?

Scatter returns a line collection and I don't know if there is a method to 
reset their x/ydata.

> For bar charts and errorbar plots, I agree ;-) How would I just blank
> the figure and replot?
I'm not sure, but maybe clf() to clear the whole figure and cla() to clear the 
axes does the job.
> (I have just been calling errorbar lots, but I'm guessing that if I add
> a legend, I'll have one entry for each time I called errorbar :-S)

regards,
Matthias

Chris Withers wrote:
>> So, basically make the x axis time instead of numbers.
>> I think the problem is actually that the daets are quite long in their
>> format. If they were rotated through 90 degress it'd likely be fine.
>> How would I do this?
> 
> I'm not sure it is the easiest way, but it works for me:
> 
> for label in ax.xaxis.get_majorticklabels():
>     label.set_rotation(+90)

To adjust tick labels when using dates, you could also try 
Figure.autofmt_xdate (I added it to the example code you were using 
below).  Internally it sets the rotation of the xticklabels as described 
above, but also sets horizontal alignment of the labels and is 
especially useful if you have multiple subplots.

-Ryan


from datetime import datetime
from time import sleep

ion()                     # interactive mode 'on'

fig = figure()

ax = fig.add_subplot(111, autoscale_on=True)

x, y = [datetime.now()], [0]
line = plot(x, y, label="my_data")[0]
# get the line-object as the first element
# of the tuple returned by plot legend()
for i in arange(30):
     x.append(datetime.now())           # append new values
     y.append(i**2)
     line.set_data(x,y)    # reset data
     ax.relim()            # reset axes limits
     ax.autoscale_view()   # rescale axes

     fig.autofmt_xdate()   # adjust the xtick labels

     draw()                # redraw current figure
     sleep(0.3)            # wait 0.3 seconds

ioff()
show()

On Fri, Mar 28, 2008 at 8:20 AM, Matthias Michler
<MatthiasMichler@...> wrote:
> On Friday 28 March 2008 13:57, Chris Withers wrote:
>  > Matthias Michler wrote:
>  > > I'm not sure it is the easiest way, but it works for me:
>  > >
>  > > for label in ax.xaxis.get_majorticklabels():
>  > >     label.set_rotation(+90)
>  >
>  > Yes, that's what I was using, just wondered if there was a better way...
>
>  At least I don't know a better way, but I'm not an expert.
>
>
>  > >> Also, how would I get this kind of updating with bar charts or
>  > >> errorbars?
>  > >
>  > > In the case of bar charts and errorbars it is quite difficult to reset
>  > > the data.
>  >
>  > Oh, I also meant to ask about scatter, can the data be easilly reset there?
>
>  Scatter returns a line collection and I don't know if there is a method to
>  reset their x/ydata.

We do not have good built in support for this kind of thing (though we
should add it).  One approach is to write a custom artist, as in this
example.  I'm using GTK only for the idle handling callback, but you
can use whatever approach works for you.  The important part is the
example showing how to write a custom artist for dynamic data:

import gtk
import numpy as np

import matplotlib
matplotlib.use('GTKAgg')
import matplotlib.pyplot as plt
import matplotlib.artist as artist
import matplotlib.colors as colors
import matplotlib.agg as agg

class DynamicMarkers(artist.Artist):
    def __init__(self, buffersize=30):
        artist.Artist.__init__(self)
        self.buffersize = buffersize
        self.x = []
        self.y = []
        self.count = 0
        self.path = None
        self.markersize = 10.
        self.facecolor = colors.colorConverter.to_rgb('blue')
        self.edgecolor = colors.colorConverter.to_rgb('black')

    def add(self, x, y):
        self.count+=1
        self.x.append(x)
        self.y.append(y)
        if self.count>self.buffersize:
            del self.x[0]
            del self.y[0]

    def draw(self, renderer):
        if self.axes is None:
            raise RuntimeError('you must first add me to the axes')

        if self.path is None:
            # use square markers
            side = renderer.points_to_pixels(self.markersize)
            offset = side*0.5

            path = agg.path_storage()
            path.move_to(-offset, -offset)
            path.line_to(-offset, offset)
            path.line_to(offset, offset)
            path.line_to(offset, -offset)
            path.end_poly()
            self.path = path

        gc = renderer.new_gc()
        self._set_gc_clip(gc)  # Artist method
        gc.set_foreground(self.edgecolor)

        renderer.draw_markers(gc, self.path, self.facecolor, self.x,
self.y, self.get_transform())


fig = plt.figure()
ax = fig.add_subplot(111)
myline = DynamicMarkers(30)
ax.add_artist(myline)
ax.set_xlim(-20, 1)

ax.set_ylim(0,1)

def animate(*args):

    i = animate.i
    print 'animate', i
    myline.add(i, np.random.rand())
    ax.set_xlim(i-30, i+1)
    fig.canvas.draw()
    animate.i += 1

    if animate.i<200: return True
    else: return False


gtk.idle_add(animate)
animate.i = 0


plt.show()

You can use the scipy version:
|~|[10]>from scipy.stats import stats
|~|[11]>stats.scoreatpercentile(x,50)
Out [11]:7.5

Le vendredi 28 mars 2008, David Simpson a écrit :
> I would like to find percentiles, with interpolation where needed, but
> the matplotlib prctile seems to be different to matlab in this respect:
>
> In [1]: x = array([ 3.0, 5.0, 7.0, 8.0, 9.0, 11.0 ])
>
> In [2]: median(x)
> Out[2]: 7.5
>
> In [3]: prctile(x,50)
> Out[3]: 8.0
>
>
> is there a function available which does include interpolation, or
> should I just write my own? (I'd also like 10th and 90th percentiles for
> example).
>
> Thanks, Dave
>
> -------------------------------------------------------------------------
> Check out the new SourceForge.net Marketplace.
> It's the best place to buy or sell services for
> just about anything Open Source.
> http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplac
>e _______________________________________________
> Matplotlib-users mailing list
> Matplotlib-users@...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users



-- 
Lionel Roubeyrie - lroubeyrie@...
Chargé d'études et de maintenance
LIMAIR - la Surveillance de l'Air en Limousin
http://www.limair.asso.fr

I guess you could read the rcParams value that corresponds to the
screen display and set rcParams['savefig.dpi'] to that value (this
might work):

mydpi = rcParams['figure.dpi']
rcParams['savefig.dpi'] = mydpi

But that seems slightly hackish and maybe not much more elegant than
what you are currently doing with passing dpi into savefig.

FWIW,

Ryan

On Thu, Mar 27, 2008 at 1:57 PM, Christopher Barker
<Chris.Barker@...> wrote:
> Ryan Krauss wrote:
>  > I think this line in the rc file is the trick
>  >
>  > #savefig.dpi       : 100
>
>  nope. I think all that does is set the default dpi for savefig. I don't
>  want any default, I want it to use the same dpi that is being used for
>  display, and I don't know ahead of time what that is, I expect it is
>  system dependent.
>
>  Or do you mean that if I comment out that line, it will default to the
>  one already in use? In which case, I need to go figure out how to change
>  that in code, rather than messing with my matplotlibrc.
>
>
>
>  -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
>
>  Chris.Barker@...
>
>  -------------------------------------------------------------------------
>  Check out the new SourceForge.net Marketplace.
>  It's the best place to buy or sell services for
>  just about anything Open Source.
>  http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
>  _______________________________________________
>  Matplotlib-users mailing list
>  Matplotlib-users@...
>  https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

Christopher Barker wrote:
> Michael Droettboom wrote:
>   
>> I think the real reason this wasn't done is that its tricky to do at the 
>> C level in a cross-platform way.  At present it uses the regular POSIX 
>> fopen in C, which isn't really "Unicode aware".
>>     
>
> The actual error is from trying to put the filename in a std::string, 
>   
Presumably because it doesn't decode into ASCII.
> but yes, I'm sure the fopen issue is the driver. Does C++ offer anything 
> better?
>   
Not in the C++ standard library in a cross platform way, that I'm aware 
of.  All three platforms (if you include Gtk, Qt etc.) have special 
Unicode-aware APIs, and each deals with things differently.  The 
approach I took basically leverages all the trouble the Python core guys 
put into to make this work cross-platform.
>    See the "Unicode
>   
>> filenames" section of the link below for some of the complications.  
>> Linux is particularly hard to get right:
>>
>> http://www.amk.ca/python/howto/unicode
>>     
>
> Thanks, that's a good one.
>
>  >  you can do the following as a workaround
>   
>> (with a performance hit from making many Python function calls):
>>
>>   savefig(open(u"CrazyUnicodeFilename.png", "w"))
>>     
>
> thanks, I'll give that a try. I'm confused, though, why the many Python 
> function calls? the C++ code doesn't just grab the file pointer?
>   
No -- a while back, it was segfaulting when trying to do that (but I've 
since figured out why -- it was a reference counting problem).  In 
0.91.2, it calls the "write" method on the Python object for each block 
of PNG data.  SVN still supports that approach (to handle arbitrary 
Python file-like objects), but will extract the real C file handle if it 
is in fact a real file.

Mike

-- 
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA