matplotlib-users Mailing List for matplotlib (Page 5)

Hi everybody,
Can you help me about two problems?
1. how to draw an arrow
----------------------------------
i'm looking for drawing single arrow (not an arrow field with quiver).
i saw there is an arrow class in the module matplotlib.patches.
So, i create an instance of the class like this:
a= Arrow(0,0,5,5,2.0)
The problem is i don't know how to draw the arrow on a figure.
My goas is to draw several arrows. Each origin of arrows is placed on the x axis. (i.e y=0)
The head of the arrow vary on x and y axes.
Here is an example:
^
/
/
/-----------------> x
 \
 \
 \ /
2. how to draw a circle?
----------------------------------
i'm trying to generate the smith chart which is very usefull in electromagnetic domain.
In this purpose, i need to draw circle.
I try to find in the polar function the way circles are drawn but i didn't find it.
Can you told me John how did you draw circles in the polar figure?
Cheers,
Philippe Collet 

Is it currently possible to make contour plots using unstructured data 
with matplotlib?
By unstructured data I mean the typical data from a 2D finite element 
simulation that
uses an unstructured polygonal mesh.
As far as I can tell the "pylab.contourf(X, Y, Z)" command requires 
structured data with X, Y
and Z each having to be 2D gridded arrays.
I have previously used the gist package to do this type of plotting 
using the
"gist.plfp(Z, X, Y, F)" command. This command allows plotting of 
polygonal data
specified by arrays Z, X, Y, F.
 - Z is the data value, shape(Z) = (NPolygons)
 - X is the vertex x coordinates, shape(X) = (NPolygons, NVertices).
 - Y is vertex y coordinates, shape(Y) = (NPolygons, NVertices)
 - F is the number of vertices for each cell, shape(F) = (NPolygons)
Is there a similar functionality in matplotlib?
Thanks
-------------------------------------
Daniel Wheeler

On 2005年4月21日 08:27:41 -0700 (PDT), Michael Brady
<mb...@jp...> wrote :
> Hi Olive,
> 
> I've had similar problems on both Solaris 8 and 9. By running python 
> under truss:
> 
> truss python
> 
> I saw that the "illegal instruction" and core dump was happening in
> the freetype libraries. I had built freetype 2.1.9 myself. When I
> deleted my own libfreetype.* and instead used the prebuilt versions 
> /usr/sfw/lib/libfreetype.* my crashes went away.
> 
> (Note that /usr/sfw/lib is the Solaris 9 location. I'm not sure if
> it's the Solaris 8 location or not).
> 
> I had built libfreetype.6.3.7 (freetype 2.1.9) and /usr/sfw/lib
> contained the earlier version libfreetype.6.3.1 . I don't whether this
> is a Solaris-only bug in later versions of freetype, or if the
> /usr/sfw/lib versions were built a different way than I built mine.
> 
> Anyway, that's how I solved the problem.
> 
> I would suggest that you try:
> 
> (1) Finding the Sun-supplied libfreetype.so binaries and using those,
> or(2) building the version of libfreetype that produces
> libfreetype.6.3.1.
> 
> Please let me know how you solve the problem.
> 
> Good luck!
> 
Hello Michael,
Like I said I solved temporarily the problem by editing the file
ft2font.cpp (from matplotlib !) at line 354 :
const char* ps_name = FT_Get_Postscript_Name(face );
I replaced it by :
const char* ps_name = NULL;
It's just a quick and dirty hack while waiting for a clean patch from
either matplotlib or freetype.
I still don't know wether it's a bug from matplotlib or freetype,
because I don't know what is the argument 'face' and what is inside...
I don't understand a thing in C++ so I can't be of much help here...
But I was working with freetype 2.1.7, I have also tried with 2.1.9 and
I also have a core dumped but no at the same place !
Don't know if my hack could work also with 2.1.9 ?
I can test if you're interested.
P.S. : I couldn't help noticing your email address, I'm also working on
space rockets but in France and also trying to promote matplotlib here !
;-)

I am trying to scale the y axis like this:
a = axes(gca())
a.set_ylim((-1.25, 1.25))
This works fine when yticks is non-empty. However, if I do
a.set_yticks([])
either before or after the set_ylim call, the y axis is not scaled.
I also tried to use
a.yaxis.set_major_locator(NullLocator())
instead of calling set_yticks([]), but to no avail.
Any suggestions are appreciated. 
Grig

Hi Olive,
I've had similar problems on both Solaris 8 and 9. By running python 
under truss:
 truss python
I saw that the "illegal instruction" and core dump was happening in the 
freetype libraries. I had built freetype 2.1.9 myself. When I deleted my 
own libfreetype.* and instead used the prebuilt versions 
/usr/sfw/lib/libfreetype.* my crashes went away.
(Note that /usr/sfw/lib is the Solaris 9 location. I'm not sure if it's 
the Solaris 8 location or not).
I had built libfreetype.6.3.7 (freetype 2.1.9) and /usr/sfw/lib contained 
the earlier version libfreetype.6.3.1 . I don't whether this is a 
Solaris-only bug in later versions of freetype, or if the /usr/sfw/lib 
versions were built a different way than I built mine.
Anyway, that's how I solved the problem.
I would suggest that you try:
(1) Finding the Sun-supplied libfreetype.so binaries and using those, or
(2) building the version of libfreetype that produces libfreetype.6.3.1.
Please let me know how you solve the problem.
Good luck!
--Michael Brady

Hello, I have been using Matplotlib lately for some
academic research. It seem to work just fine. The only
serious problem is a problem with memory. When I run
the following code, I get the memory increasing
linearly with the time. Soon (after around 20
generations, and 100 times is this part of the program
 called) I get beyond the limits of my laptop, and the
simulation ends. Considering that I need to do at
least 200 generations, you can see that it is a
serious problem.
If I comment the savefig line the memory does not
grow.
I tried to add a garbage collection but I got no
result.
Data is a list of list of data: i.e.
[[1,2,3][2,7,8][32,45,67]]
symbol2color a dictionary that for each list gives a
different color.
def fillstrip(lower,upper,color):
 reppu=upper[:]
 reppu.reverse()
 y=lower+reppu
 x1 = range(0, len(lower), 1)
 x2 = x1[:]
 x2.reverse()
 x=x1+x2
 fill(x, y, color)
def area(data,color):
 "like the area matlab function"
 lower=[0]*len(data[0])
 for (upper,c) in zip(data,color):
 upper=map((lambda x,y:x+y),lower,upper)
 fillstrip(lower,upper,c)
 lower=upper
def PlotAreaOrganization(data,filename):
 keys=symbol2color.keys()
 keys.sort()
 colors=[]
 for key in keys: colors.append(symbol2color[key])
 fig=figure(dpi=75)
 area(data,colors)
 savefig(filename)
 close(fig)
Any suggestion? Is there any other way to save a
picture that might work better?
Pietro
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

Hi Gabriele,
Matplotlib currently does not do this. The problem, as I understand it (and 
there are others on this list that may correct me), is that there is a lot of 
information that needs to be written into the prologue at the beginning of 
the file, like font information. This makes it difficult to append to an 
existing file, since each page description could require different 
information in the prologue.
As an alternative, you could create multiple eps files. In the process, you 
could write a script that generates a simple latex document and embeds your 
eps files. Then you could convert that into PS. There are examples of this in 
"Python Scripting for Computational Science" by Hans Petter Langtangen.
Darren
On Wednesday 20 April 2005 3:29 am, Gabriele Garavini wrote:
> Hi,
> I'm new to matplotlib, I'm trying to save a series of plots in a single
> postscript files. Is that possible?
> I'm able to save each single plot in a different file using savefig,
> but I'd like my script to save the plots in a single ps file with more
> pages. Is there anyway to open a ps file, send the output of the plot()
> command to it and close it as the script ends?
>
> Thanx
> Cheers
> Gabriele
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: New Crystal Reports XI.
> Version 11 adds new functionality designed to reduce time involved in
> creating, integrating, and deploying reporting solutions. Free runtime
> info, new features, or free trial, at:
> http://www.businessobjects.com/devxi/728
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
-- 
Darren S. Dale
Bard Hall
Department of Materials Science and Engineering
Cornell University
Ithaca, NY. 14850
dd...@co...

I would like to use matplotlib to make a simple client-side
image map for a web page. I have a scatter plot and I would
like mouseovers/clicks to show information about the given
point.
I started with the webapp_demo.py from the most recent
release (0.80). There was a small bug
% python webapp_demo.py
Traceback (most recent call last):
 File "webapp_demo.py", line 87, in ?
 make_fig()
 File "webapp_demo.py", line 64, in make_fig
 x = nx.rand(100)
AttributeError: 'module' object has no attribute 'rand'
Here's my bandage hack
import random
def rand(n):
 return nx.array([random.random() for i in range(n)])
nx.rand = rand
I then added some code after the PNG is written which
generates an basic html file from the coordinates using
what I thought was the correct transformation from
data space to image space.
 canvas.print_figure('webapp.png', dpi=150)
 f = open("webapp.html", "w")
 f.write('''<HTML><BODY><img src="webapp.png" ismap usemap="#plot">
<map name="plot">
''')
 t = c.get_transform()
 xys = t.seq_xy_tups(c.get_verts())
 for i, xy in enumerate(xys):
 f.write('<area shape="circle" coords="%d,%d,2" href="%d">\n' %
 (xy[0], xy[1], i))
 f.write("</map>\n</html>")
The HTML output starts with the following
<HTML><BODY><img src="webapp.png" ismap usemap="#plot">
<map name="plot">
<area shape="circle" coords="393,286,2" href="0">
<area shape="circle" coords="381,285,2" href="1">
<area shape="circle" coords="222,271,2" href="2">
<area shape="circle" coords="238,300,2" href="3">
There are two things wrong with the code. First, the
transformed coordinates aren't at the same scale as
the generated PNG. The PNG size depends on the dpi
setting but the x/y coordinates aren't affected by
that setting.
Second, I think I have the y direction swapped. Where
do I get the current image size, and for that matter
how do I set the output image size?
					Andrew
					da...@da...

Hi All,
I think I found a solution for dates.py which could work for everyone.
Instead of hardcoding the encoding I "import locale" and get the default 
encoding from it. I have tested it on both Win XP and 2000 and it works 
for me, i.e. in English locale I get "Dec" and in French locale I get 
"déc." etc. without the correction I would get a graphic sign for the "é".
To test just use "locale.setlocale(locale.LC_ALL, 'fr')"
Just noticed that "avr" is not correctly aligned with the other month, 
but that's for another day.
The only two changes needed to dates.py are:
import locale
and change
return dt.strftime(fmt)
to
return unicode(dt.strftime(fmt), locale.getpreferredencoding())
Best regards
Werner
Werner F. Bruhin wrote:
> In some of my plots I show month names as axes labels.
> 
> When I run this on a machine in "French" (i.e. on Win XP changing the 
> Settings - Regional and Language settings" to French the labels don't 
> show correctly if there is an accented character in there e.g. "déc".
> 
> I changed the line 256 in dates.py (matplotlib 0.8) from:
> 
> return dt.strftime(fmt)
> 
> to:
> 
> return unicode(dt.strftime(fmt), 'iso-8859-1')
> 
> Obviously this is not a correct fix as it will only work in some 
> situations.
> 
> I tried to use sys.getdefaultencoding(), instead of hard coding 
> 'iso-8859-1', but on my machine it returns "ascii".
> 
> Any suggestions on how to handle this correctly will be very appreciated.
> 
> See you
> Werner
> 
> 
> 
> -------------------------------------------------------
> SF email is sponsored by - The IT Product Guide
> Read honest & candid reviews on hundreds of IT Products from real users.
> Discover which products truly live up to the hype. Start reading now.
> http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
> 

Hi,
I'm new to matplotlib, I'm trying to save a series of plots in a single 
postscript files. Is that possible?
I'm able to save each single plot in a different file using savefig, 
but I'd like my script to save the plots in a single ps file with more 
pages. Is there anyway to open a ps file, send the output of the plot() 
command to it and close it as the script ends?
Thanx
Cheers
Gabriele

Hi,
I'm new to matplotlib, I'm trying to save a series of plots in a single 
postscript files. Is that possible?
I'm able to save each single plot in a different file using savefig, 
but I'd like my script to save the plots in a single ps file with more 
pages. Is there anyway to open a ps file, send the output of the plot() 
command to it and close it as the script ends?
Thanx
Cheers
Gabriele

Hi,
I'd like to be able to add text to various (x,y) coordinates. Currently
I specify a figsize of (1.3,0.3) and I add as text (via the text
method) the value of the last y, right next to that data point.
Depending on the actual value, sometimes the text does not fit within
the bounds of the figure. 
Is there any way I can enlarge just the figure so that the text can
fit, without enlarging the axes? Or is there a better way to solve my
problem?
Thanks,
Grig

On Friday 15 April 2005 5:32 am, Sascha Schnepp wrote:
> I realized a change in the behavior of the eps export with the release 0.8.
> Sometimes (not always!) the bounding box on the left side is very close to
> the ylabel. When converting this to a pdf and printing it, this results in
> a (partially or even totally) cutted ylabel on your printout...
The script at the end of this message will reproduce the behavior Sasha 
reported.
Sasha, the short answer is to try replacing your call pl=subplot(111) with 
pl = axes([0.16,0.11,0.8,0.8]).
The problem is that MPL will not be able to guess the appropriate position of 
the axes in every circumstance, because the precision of the tick labels is 
allowed to go as far as 3 or 4 places, and the presence scientific notation 
will also push the axis label off the page.
I did some work on a new formatter a while back, and think it would help to 
solve this issue more generally. The idea was to pull the scientific notation 
out of the ticklabels, and render it at the top of the axis, like Matlab 
does. Another place to render could be in the axis label, like Igor allows. 
It might also be appropriate to add a new rc setting for the tick precision: 
the number of sigfigs to include in a ticklabel. MPL would then be able to 
intelligently place the axis in the plot window, based on the rc settings for 
fontsize of the axis labels, the fontsize of the ticklabels, and the 
precision.
from pylab import *
a=arange(-1,1,.01)
plot(a,a*1e-4)
xlabel('X axis')
ylabel('Y axis')
savefig('test.eps')
show()
-- 
Darren S. Dale
Bard Hall
Department of Materials Science and Engineering
Cornell University
Ithaca, NY. 14850
dd...@co...

Hi Nikolai,
Nikolai Hlubek wrote:
> Last week I stumbled about a bug in imshow, which can easily be
> reproduced by doing the following:
>
> python
> >>> from pylab import *
> >>> imshow(ones((20,40)),aspect='preserve',origin='lower')
> >>> show()
> Note that the y-axis is wrong and resizing makes this even worse.
> I tried versions 0.71 to 0.80 which show all the same behaviour.
>
> It is obviously the aspect='preserve' option, that introduces this bug.
>
> So I'd like to address two issues:
> First of course I'd like to ask if this bug could be fixed since
> constant aspect ratio is very crucial for me.
> Second it seems that the plotting area is always maximum size and the
> aspect='preserve' option just rescales the axis. I think it would be
> more natural to rescale the plotting area.
You may want to use matshow for this kind of problem, which will try to give you
a figure with the aspect ratio of the array you actually want to display.
As you noticed, the preserve feature is basically broken right now, and there
has been a fair amount of discussion on how to fix it. In the meantime, the
best workaround is to compute figure size and axes aspect rations appropriate
for your image aspect ration using the figure and axes command, which is
basically what matshow does.
Best,
f/jdh

There appears to be a bug in the latest basemap distrib. (0.3.1) that
prevents it from being installed in a non-standard location. 
In basemap.py, the required data file location is specified as:
_datadir =
os.path.join(sys.prefix,'share/basemap-py'+repr(sys.version_info[0])+rep
r(sys.version_info[1]))
On my system, sys.prefix returns '/usr'. However, I installed the
package using the --prefix option and a different path than /usr. Then
when basemap tries to access data files for coastlines, etc. with
open(os.path.join(_datadir,'gshhs_'+resolution+'.txt'))
it's looking for the files in /usr and they ain't there.
Since I don't have root privileges on my own machine, I can't do a
symbolic link (and that would be tacky anyway). Jeff, could you fix
this? Thanks.
Jeff Orrey

On 2005年4月16日 00:19:37 +0200, Olive <enc...@ya...> wrote
:
> I forgot to take the version number of the freetype library before
> leaving work, I will get it on monday.
> 
Version of freetype is 2.1.7 !

Hi,
The more I am using matplotlib, the more I like it. But using it more 
means I more often hit the limits of the code, so I was wondering if 
there is any plan for a routine to plot data points w.r.t 3 axes (so a 
perspective scatter plot with x,y,z, data and a viewing angle). Of 
course another (maybe more difficult) goal would be to be able to draw 
surfaces, but already a simple scatter plot in 3D would help 
tremendously there (I realize the concept of "axis" has then to be 
revised, but well I have no idea how much effort this is).
(if there is no - coming - plan for this, is there any way around it?)
Please let me know, and congrats for the work already achieved!
Cheers
Eric
-- 
===============================================================
Observatoire de Lyon ems...@ob...
9 av. Charles-Andre tel: +33 4 78 86 83 84
69561 Saint-Genis Laval Cedex fax: +33 4 78 86 83 86
France http://www-obs.univ-lyon1.fr/eric.emsellem
===============================================================

Hi everyone
Last week I stumbled about a bug in imshow, which can easily be 
reproduced by doing the following:
python
 >>> from pylab import *
 >>> imshow(ones((20,40)),aspect='preserve',origin='lower')
 >>> show()
Note that the y-axis is wrong and resizing makes this even worse.
I tried versions 0.71 to 0.80 which show all the same behaviour.
It is obviously the aspect='preserve' option, that introduces this bug.
So I'd like to address two issues:
First of course I'd like to ask if this bug could be fixed since 
constant aspect ratio is very crucial for me.
Second it seems that the plotting area is always maximum size and the 
aspect='preserve' option just rescales the axis. I think it would be 
more natural to rescale the plotting area.
Cheers,
Nikolai

Hi David,
Thanks for the info.
See you
Werner
D Brown wrote:
> Werner,
> 
> Recently I reported a batch plot memory issue when using
> mostly matlab-style commands and TkAgg backend. In my case
> I found that the memory increased when I called close(). 
> If I stopped using this the memory use became more or less
> constant. Based on Fernando Perez's comments I also
> started using import gc and gc.collect() to reduce memory
> use more. This just calls the garbage collector more
> frequently to clean up memory. If your system is paging to
> disk because of memory use it can speed things up quite a
> bit. In my case it keeps the total python footprint
> significantly reduced.
> 
> Good Luck,
> 
> David
> 
> 
>>From: "Werner F. Bruhin" <wer...@fr...>
>>Date: 2005年4月14日 11:42:34 +0200
>>Subject: [Matplotlib-users] Memory usage
> 
> 
>>Hi All,
>>
>>Doing multiple plots I see that memory usage grows and
> 
> grows.
> 
>>Before doing a new plot I do "self.figure.clear()", or 
>>"self.figure.clf()", is there some other call I should
> 
> use?
> 
>>See you
>>Werner
> 
> 
> +_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+
> 
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around 
> http://mail.yahoo.com 
> 
> 
> -------------------------------------------------------
> SF email is sponsored by - The IT Product Guide
> Read honest & candid reviews on hundreds of IT Products from real users.
> Discover which products truly live up to the hype. Start reading now.
> http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click

In some of my plots I show month names as axes labels.
When I run this on a machine in "French" (i.e. on Win XP changing the 
Settings - Regional and Language settings" to French the labels don't 
show correctly if there is an accented character in there e.g. "déc".
I changed the line 256 in dates.py (matplotlib 0.8) from:
return dt.strftime(fmt)
to:
return unicode(dt.strftime(fmt), 'iso-8859-1')
Obviously this is not a correct fix as it will only work in some situations.
I tried to use sys.getdefaultencoding(), instead of hard coding 
'iso-8859-1', but on my machine it returns "ascii".
Any suggestions on how to handle this correctly will be very appreciated.
See you
Werner

Hi Stephen,
On Fri, Apr 15, 2005 at 10:01:30AM -0700, Stephen Walton wrote:
> What I really meant was that the PostScript file produced by MPL starts=
=20
> with '%!PS-Adobe-3.0', which I thought meant Postscript level 3. =20
> Changing the 3.0 to 2.0 by hand still doesn't allow it to print. dvips=
=20
> output files print fine on this printer, and they start with=20
> '%!PS-Adobe-2.0'.
Confusingly enough there are two kinds of version numbers involved, here.
There is the PostScript language level, which can be 1, 2, or 3.
Additionally there is something called DSC (Document structuring
conventions), which gives additional information in PostScript comments
(lines starting with a percent sign). This DSC thing allows PostScript
viewers to go back to the previous page, allows to print single pages
=66rom a PostScript file, etc. DSC also comes in several versions.
The "%!PS-Adobe-3.0" indicated that the document confirms to DSC version 3.
This does not imply anything for the PostScript language level. Since
all DSC stuff is most probably ignored by the Printer, it should not
cause any problems.
> Thanks for the suggestion, but I'm afraid this didn't help. The=20
> workaround I found was to do a dvipdf on the LaTeX file which included=20
> the MPL doc, read it with Acrobat Reader, and tell acroread to print it=
=20
> at Postscript level 1.
GhostScript has also a converting-fancy-PostScript-into-simple-PostScript
functionality which might come in handy here.
All the best,
Jochen
--=20
http://seehuhn.de/

Werner,
Recently I reported a batch plot memory issue when using
mostly matlab-style commands and TkAgg backend. In my case
I found that the memory increased when I called close(). 
If I stopped using this the memory use became more or less
constant. Based on Fernando Perez's comments I also
started using import gc and gc.collect() to reduce memory
use more. This just calls the garbage collector more
frequently to clean up memory. If your system is paging to
disk because of memory use it can speed things up quite a
bit. In my case it keeps the total python footprint
significantly reduced.
Good Luck,
David
>From: "Werner F. Bruhin" <wer...@fr...>
>Date: 2005年4月14日 11:42:34 +0200
>Subject: [Matplotlib-users] Memory usage
>Hi All,
>
>Doing multiple plots I see that memory usage grows and
grows.
>
>Before doing a new plot I do "self.figure.clear()", or 
>"self.figure.clf()", is there some other call I should
use?
>
>See you
>Werner
+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

The attached backend_wx (matplotlib 0.8) has a Printer_Setup2 method 
using the standard wx page setup stuff. There seems to have been 
problems with this in the past that is why I created a new method and 
left Printer_Setup alone.
I also commented out "self.printerData.SetPaperId(wx.PAPER_LETTER)"
in Printer_Init, not setting it will use the printer default paper size 
which I think is more correct, so if I have a nice A3 capable printer I 
don't have to reset my paper size all the time, or if I live e.g. in the 
South of France I can use my standard A4 paper.
See you
Werner

On 2005年4月06日, John Hunter apparently wrote: 
> >>> rect.set_facecolor(0.8) # a grayscale intensity 
...
> There are many wanys to set colors in matplotlib, these are discussed 
> at http://matplotlib.sourceforge.net/matplotlib.pylab.html#-colors 
I do not see any discussion of grayscale at that location.
Grayscale does work mostly as expected once one knows about 
it.
Exception: integer values are not welcome.
(E.g., pylab tries to treat 0 as part of an RGB 
specification.)
Cheers,
Alan Isaac

John Hunter wrote:
>See examples/shared_axis_demo.py,
>examples/two_scales.py and examples/ganged_plots.py.
>
Speaking of which: I just put a ganged plot made with matplotlib into a 
grant proposal. It worked really well with one exception: the top 
value of the y axis on each graph was printed on top of the bottom value 
of the graph above. If you look carefully, you'll see that 
examples/ganged_plots.py does the same thing.
Any good way to avoid this?