matplotlib-users Mailing List for matplotlib (Page 6)

Hello,
I can successfully change the facecolor in a figure but when I try and save it to a file (jpg or png) the color goes away - the rest of the plot remains only the facecolor reverts to white. I have tried using GTK and GTKAgg. I have also tried changing the 'savefig' value in the RC file - though I'm not sure this is relevant because I'm not using the savefig command. So far nothing is working. Any ideas?
Thanks,
Mike
 
---------------------------------
Catch up on fall's hot new shows on Yahoo! TV. Watch previews, get listings, and more!

Hello,
I noticed, that bar() with log=3DTrue plots very strange graphs. In fact,
the bars in this case grow from the bottom of the graph (I guess from
the value of log(+0), i.e. -=E2=88=9E). This way the relative height of the=
 bars
says almost nothing about the value of data, because the bars are
higher, the lower is ylim()[0]. It is hard to distinguish data values
above and below 1 (positive and negative log).
I think that in many situations it is more useful to base bars on
the level of log(y)=3D0. This is achievable with manual log-scaling of
data, yet in this case it also requires manual tuning of ylabels.
I attach a script and an image which show the default plotting in
comparison to grow-from-log(1) plotting to see the difference, and
suggest making it possible to change the level from which the bars grow.
Best regards,
jetxee

Christian Meesters <mee...@un...>
writes:
> is it somehow possible to have a hatch in parts of the background, which
> would achieve something like this pseudo-parameter to axvspan
> pylab.axvspan(2, 10, hatch='//')?
Do you mean something like this?
In [34]: phi=pi*array((0,.2,.4,.6,.8,1,-.8,-.6,-.4,-.2))
In [35]: fill(cos(phi), sin(phi))
Out[35]: [<matplotlib.patches.Polygon instance at 0x1894cda0>]
In [36]: a=gca()
In [37]: setp(getp(a,'frame'), hatch='//')
Out[37]: [None]
(For some reason I don't see the hatch pattern in Agg-based backends, in
current svn, but it is there in e.g. eps and pdf.)
-- 
Jouni K. Seppänen
http://www.iki.fi/jks

James Boyle <bo...@ll...> writes:
> I have not been able to figure out how to just make the first and 
> last ytick labels vanish. [...]
> I thought that the following might work but this just makes all the 
> labels disappear - my understanding is incomplete.
> ytl = a.get_yticklabels()
> ytl[0]._visible = False
> ytl[-1]._text = False
It is usually a bad idea to manipulate directly anything starting with
an underscore -- that's a Pythonic way of indicating a "private"
variable. The set_visible() method should work here:
ytl = a.get_yticklabels()
ytl[0].set_visible(False)
-- 
Jouni K. Seppänen
http://www.iki.fi/jks

> You want legend((bar1[0],bar2[0]), ('First','Second')). What happened
> was that matplotlib made a legend entry for two of the blue bars in
> bar1; it would have made six entries, but stopped because you only gave
> it two labels.
>
> 
Dear Jouni,
thanks for all the answers.
Gianluca

Gianluca Santarossa
<gia...@ch...> writes:
> In this example, on my PC both the entries in the legend appear in blue 
> color:
> legend((bar1,bar2), ('First','Second'))
You want legend((bar1[0],bar2[0]), ('First','Second')). What happened
was that matplotlib made a legend entry for two of the blue bars in
bar1; it would have made six entries, but stopped because you only gave
it two labels.
> Moreover, if I add a legend to a graph plotting a set with marks and 
> without lines, the legend will show two points instead of one (which 
> would have been the expected behaviour). Is this correct?
I get four points, not two, but perhaps this has changed in the svn
version. At least in the svn version you can control the number of
points with the numpoints keyword argument:
In [6]: plot([3,1,4,1,5,9,2,6,5,3,5],'bo')
Out[6]: [<matplotlib.lines.Line2D instance at 0x16cf9b98>]
In [7]: legend(_, ('foo',), numpoints=1)
Out[7]: <matplotlib.legend.Legend instance at 0x16cf9bc0>
-- 
Jouni K. Seppänen
http://www.iki.fi/jks

Dear all,
I am an unexperienced matplotlib user, and I have a couple of questions 
about adding a legend to a graph.
In this example, on my PC both the entries in the legend appear in blue 
color:
#!/usr/bin/env python
import numarray as na
from pylab import *
labels = ["A", "B", "C"]
first = [ 1, 2, 3 ]
second = [ 3, 2, 1 ]
xlocations = na.array(range(len(labels)))+0.33
width = 0.33
bar1=bar(xlocations, first, width=width, color='blue')
bar2=bar(xlocations+.33, second, width=width, color='red')
legend((bar1,bar2), ('First','Second'))
show()
Did I make some mistake?
Moreover, if I add a legend to a graph plotting a set with marks and 
without lines, the legend will show two points instead of one (which 
would have been the expected behaviour). Is this correct?
Thank in advance,
Gianluca

On 13/09/2007, Alan G Isaac <ai...@am...> wrote:
> http://matplotlib.sourceforge.net/matplotlib.pylab.html#-plot_date
Thanks, that's helpful and has given me a push in the right direction.
I've got a rough code producing the plot I need - just need to clean
it up now.
Cheers
Adam

Also in the examples, there is date_demo1.py...
(And others.)
Cheers,
Alan Isaac

http://matplotlib.sourceforge.net/matplotlib.pylab.html#-plot_date
hth,
Alan Isaac

Hi
I need to produce a line plot of some data against the date in ISO
format, i.e. the data is something like:
20060412 546
20060413 547
20060414 657
20060415 438
...
I've been looking at the examples and can't find anything appropriate.
 As far as I can tell from the documentation I need to read in the
date and use date2num to convert the date into matplotlibs internal
format for representing dates... but I'm not sure - is this the
correct approach?
Is there any example code that I'm missing that plots data against the date?
Cheers
Adam

Hi,
I have a histogram with orders of magnitude difference in counts of each
bin. I want to use a log yscale in plotting it. But there are bins with 0
counts. What's the best way to plot it? I've read the log_bar.py example
which uses bar() for plotting. It works when I pull it into a script. But
I'd appreciate any pointers to better ways.
Thanks,
Ping

On 2007年9月13日, Fabian Braennstroem apparently wrote:
> Does nobody have an idea; especially for the 'dynamic'
> number of plotted arrays!? 
The question is unclear.
The problem seems easy enough,
if you get your hands on the arguments to your script.
http://homepage.mac.com/andykopra/pdm/tutorials/simplifying_script_arguments.html
hth,
Alan Isaac

Does nobody have an idea; especially for the 'dynamic'
number of plotted arrays!?
Regards!
Fabian
Fabian Braennstroem schrieb am 09/09/2007 09:01 PM:
> Hi,
> 
> I have a small script which reads a csv file with several
> columns and puts it into an scipay array, which
> I can plot using matplotlib. It works fine, but just with
> explicitly setting the number of columns:
> 
> res=loglog(array_mapped[:,0],array_mapped[:,1], 'b',
> array_mapped[:,0],array_mapped[:,2], 'g',
> array_mapped[:,0],array_mapped[:,3], 'r',
> array_mapped[:,0],array_mapped[:,4], 'y',
> array_mapped[:,0],array_mapped[:,5], 'k',
> array_mapped[:,0],array_mapped[:,6], '-bo',
> linewidth = 2)
> 
> Is there a way to define the number of columns, which I want
> to plot? Or even better, can I apply an sys.argv to
> define the plotted columns, e.g.
> "python csvplot.py all": which plots all columns with its
> value for the y-coordinate and the line-number as x-coordinate
> "python csvplot.py all1": which does the same as above, but
> using column 1 as abscissae
> "python csvplot.py 1 2 5": which plots columns 2 and 5
> against column 1...
> 
> Would be nice, if anybody has an idea, how to achieve this!?
> Regards!
> Fabian
> 
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/

Hi,
pylab.scatter(x,y)
pylab.errorbar(x,y,yerr,linestyle='None')
if you want to use scatter, or alternatively:
pylab.errorbar(x,y,yerr,fmt='o',linestyle='None')
Manuel
Armen Nalian wrote:
> Hello,
> I was wondering where I can get an example of a scatter plot with error 
> bars.
> I have time measurements and the stds
> 0 23 0.23
> 1 25.1 0.21
> 4 27 0.1
> 7 29 0.21
> 9 35 0.1
> 
> 
> how can I get a scatter plot with error bars? I know I can get a bar 
> plot, but I need just a symbol with error bars.
> Thank you,
> Armen
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Hi,
pylab.scatter(x,y)
pylab.errorbar(x,y,yerr,linestyle='None')
if you want to use scatter, or alternatively:
pylab.errorbar(x,y,yerr,fmt='o',linestyle='None')
Manuel
Armen Nalian wrote:
> Hello,
> I was wondering where I can get an example of a scatter plot with error 
> bars.
> I have time measurements and the stds
> 0 23 0.23
> 1 25.1 0.21
> 4 27 0.1
> 7 29 0.21
> 9 35 0.1
> 
> 
> how can I get a scatter plot with error bars? I know I can get a bar 
> plot, but I need just a symbol with error bars.
> Thank you,
> Armen
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Hello,
I was wondering where I can get an example of a scatter plot with error 
bars.
I have time measurements and the stds
0 23 0.23
1 25.1 0.21
4 27 0.1
7 29 0.21
9 35 0.1
how can I get a scatter plot with error bars? I know I can get a bar 
plot, but I need just a symbol with error bars.
Thank you,
Armen

On 2007年9月13日, Steve Schmerler apparently wrote: 
> two_scales.py in the mpl examples? 
Yes, that is pretty good, and much better
than the approaches used in the examples
behind the screen shots:
http://matplotlib.sourceforge.net/screenshots.html
Thanks!
Alan

Alan G Isaac wrote:
> I need to make some dual y-scale plots:
> on time series plotted against the left axis,
> with a second plotted again the right axis (which has its 
> own scale). I think Matplotlib did not used to provide
> dual scale plotting: is it now available?
> 
You mean something like two_scales.py in the mpl examples? (in svn or download 
from the homepage)
-- 
cheers,
steve
Random number generation is the art of producing pure gibberish as quickly as 
possible.

I need to make some dual y-scale plots:
on time series plotted against the left axis,
with a second plotted again the right axis (which has its 
own scale). I think Matplotlib did not used to provide
dual scale plotting: is it now available?
Thank you,
Alan Isaac

Thanks Michael! I installed fondu and converted my Helvetica.dfont to .ttf
and put them in /Users/ybendana/Library/Fonts. However, the fontManager wa=
s
still picking up the /System/Library/Fonts/Helvetica.dfont. As a
workaround, I edited font_manager.py line 209 in findSystemFonts() to build
the font cache from only .ttf fonts:
 for f in OSXInstalledFonts(fontext=3Dfontext):
After deleting the ~/.matplotlib/ttfont.cache, it now picks up the
Helvetica.ttf font.
yuri
On 9/12/07, Michael Droettboom <md...@st...> wrote:
>
> Yuri,
>
> A .dfont file can be seen as a collection of .ttf files. A .dfont
> generally will contain the regular, bold and italic versions of a font,
> but matplotlib is hardcoded to only look at the first of them. So when
> you request "sans-serif", you get Helvetica. When you request
> "sans-serif:bold", it fails, (since it can't get to the bold version)
> and falls back on the default matplotlib font, Vera Sans Regular.
>
> Long term, we probably need to improve matplotlib to support dfonts
> correctly. As a workaround, you can adjust your font settings to refer
> to .ttf fonts, or convert your dfonts to ttf fonts using fondu.
>
> Cheers,
> Mike
>
> Yuri Benda=F1a wrote:
> > Hi,
> >
> > I'm a new user and I installed matplotlib 0.90.1 on Mac 10.4.9. I'm
> > using the WxAgg backend, which I installed using MacPorts. I haven't
> > had a problem so far except when I try to change font properties like
> > family, fontname, and weight. For example, the following doesn't work=
:
> >
> > figtext(0.8,0.3,"Large Folds", weight=3D'bold', fontname=3D'Helvetica')
> >
> > I only get the default font (I believe it's Vera) and no bold. This
> > also happens with the PDF backend so I don't think that's the problem.
> > I checked the fontManager object and verified that it knows where the
> > fonts are installed on my Mac. I can call findfont() and it does retur=
n
> > the Helvetica font:
> >
> > In [10]: fm.fontManager.findfont(fm.FontProperties('sans-serif'))
> > Out[10]: '/System/Library/Fonts/Helvetica.dfont'
> >
> > Does anyone have a solution for this?
> >
> > thanks,
> >
> > yuri
> >
> >
> >
> > -----------------------------------------------------------------------=
-
> >
> >
> -------------------------------------------------------------------------
> > This SF.net email is sponsored by: Microsoft
> > Defy all challenges. Microsoft(R) Visual Studio 2005.
> > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> >
> >
> > -----------------------------------------------------------------------=
-
> >
> > _______________________________________________
> > Matplotlib-users mailing list
> > Mat...@li...
> > https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
> --
> Michael Droettboom
> Operations and Engineering Division
> Space Telescope Science Institute
> Operated by AURA for NASA
>
--=20
Yuri Benda=F1a
Graduate Student
UC Berkeley/UCSF Joint Graduate Group in Bioengineering

"Yuri Bendaña" <ybe...@be...>
writes:
> For example, the following doesn't work:
> figtext(0.8,0.3,"Large Folds", weight='bold', fontname='Helvetica')
> I only get the default font (I believe it's Vera) and no bold. 
Yes, I think that's a known bug. The font manager doesn't know how to
read the bold version of Helvetica from Helvetica.dfont. A workaround is
to run fondu (http://fondu.sourceforge.net/) on Helvetica.dfont, put the
resulting ttf files in some directory and use them directly:
In [20]: from matplotlib.font_manager import FontProperties
In [21]: fp=FontProperties(fname='/tmp/HelveticaBold.ttf',size=14)
In [22]: text(.1,.1,'this is bold',fontproperties=fp)
-- 
Jouni K. Seppänen
http://www.iki.fi/jks

Yuri,
A .dfont file can be seen as a collection of .ttf files. A .dfont 
generally will contain the regular, bold and italic versions of a font, 
but matplotlib is hardcoded to only look at the first of them. So when 
you request "sans-serif", you get Helvetica. When you request 
"sans-serif:bold", it fails, (since it can't get to the bold version) 
and falls back on the default matplotlib font, Vera Sans Regular.
Long term, we probably need to improve matplotlib to support dfonts 
correctly. As a workaround, you can adjust your font settings to refer 
to .ttf fonts, or convert your dfonts to ttf fonts using fondu.
Cheers,
Mike
Yuri Bendaña wrote:
> Hi,
> 
> I'm a new user and I installed matplotlib 0.90.1 on Mac 10.4.9. I'm 
> using the WxAgg backend, which I installed using MacPorts. I haven't 
> had a problem so far except when I try to change font properties like 
> family, fontname, and weight. For example, the following doesn't work:
> 
> figtext(0.8,0.3,"Large Folds", weight='bold', fontname='Helvetica')
> 
> I only get the default font (I believe it's Vera) and no bold. This 
> also happens with the PDF backend so I don't think that's the problem. 
> I checked the fontManager object and verified that it knows where the 
> fonts are installed on my Mac. I can call findfont() and it does return 
> the Helvetica font:
> 
> In [10]: fm.fontManager.findfont(fm.FontProperties('sans-serif'))
> Out[10]: '/System/Library/Fonts/Helvetica.dfont'
> 
> Does anyone have a solution for this?
> 
> thanks,
> 
> yuri
> 
> 
> 
> ------------------------------------------------------------------------
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
-- 
Michael Droettboom
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA

On 9/12/07, Jouni K. Sepp=E4nen <jk...@ik...> wrote:
>
> Martin Bures <nee...@ya...>
> writes:
>
> > MATLAB has a command, print, and it allows you to output a figure to a
> > file, such as a .ps file and it has a switch '-append' so that you can
> > append multiple plots to the same file.
> >
> > Is there a switch for the pylab command, 'savefig' to do the same
> > thing?
>
> No, at least not currently.
>
> --
> Jouni K. Sepp=E4nen
> http://www.iki.fi/jks
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
somthing like append(filename) will be the best thing to do it.
but this will not work for *.png files.
I am not familiar with EPS and PS. I thought only ps supports
multipages, and each page has one figure.
Since my data analysis and plotting works sometimes on Linux(office),
sometimes Windows (home). A native support from matplotlib would be
fantasitic.
Thanks for the suggestions.

Hi,
I'm a new user and I installed matplotlib 0.90.1 on Mac 10.4.9. I'm using
the WxAgg backend, which I installed using MacPorts. I haven't had a
problem so far except when I try to change font properties like family,
fontname, and weight. For example, the following doesn't work:
figtext(0.8,0.3,"Large Folds", weight='bold', fontname='Helvetica')
I only get the default font (I believe it's Vera) and no bold. This also
happens with the PDF backend so I don't think that's the problem. I checked
the fontManager object and verified that it knows where the fonts are
installed on my Mac. I can call findfont() and it does return the Helvetica
font:
In [10]: fm.fontManager.findfont(fm.FontProperties('sans-serif'))
Out[10]: '/System/Library/Fonts/Helvetica.dfont'
Does anyone have a solution for this?
thanks,
yuri