matplotlib-users Mailing List for matplotlib (Page 4)

Hi,

I am trying to generate graphs using

text.usetex : True
ps.usedistiller : xpdf

Unfortunately, when running

import pylab as P
P.plot([10],[10])
P.savefig('test.eps')

I get an error message:

File "/home/stefan//lib/python2.4/site-packages/matplotlib/backends/backe=
nd_ps.py", line 1412, in get_bbox
    raise RuntimeError('Ghostscript was not able to extract a bounding bo=
x.\
RuntimeError: Ghostscript was not able to extract a bounding box.Here is =
the Ghostscript output:

ESP Ghostscript 815.02: Unrecoverable error, exit code 1

When I run with --verbose-debug-annoying I see

pdftops -paper match -level2 "/tmp/098f6bcd4621d373cade4e832627b4f6.pdf" =
"/tmp/098f6bcd4621d373cade4e832627b4f6.ps" > "/tmp/098f6bcd4621d373cade4e=
832627b4f6.output"

gs -dBATCH -dNOPAUSE -sDEVICE=3Dbbox "/tmp/098f6bcd4621d373cade4e832627b4=
f6"
ESP Ghostscript 815.02 (2006-04-19)
Copyright (C) 2004 artofcode LLC, Benicia, CA.  All rights reserved.
This software comes with NO WARRANTY: see the file PUBLIC for details.
ERROR: /undefined in pdfLastFill

I am running gs v8.15.2, pdftops v3.00, dvipng v1.5 and xpdf v3.01.  Are
these versions known to be incompatible, or is there possible a
problem with the output generated by matplotlib?

With

ps.usedistiller : ghostscript

everything runs smoothly, but the text looks pretty dismal.

I would appreciate any advice on how to get this running!

Thanks
St=E9fan

The standard plot:
 
ax.plot_date(dates,values) 
does a "dot" plot
 
whereas 
 
ax.plot_date(dates,values,'-') 
plots a line. 
 
How, though, do I plot values as bars?
 
Thanks
Derek
 
PS dates are converted from an array "dates as strings",
using the datestr2num() - which works well!


-- 
This message is subject to the CSIR's copyright, terms and conditions and
e-mail legal notice. Views expressed herein do not necessarily represent the
views of the CSIR.
 
CSIR E-mail Legal Notice
http://mail.csir.co.za/CSIR_eMail_Legal_Notice.html 
 
CSIR Copyright, Terms and Conditions
http://mail.csir.co.za/CSIR_Copyright.html 
 
For electronic copies of the CSIR Copyright, Terms and Conditions and the CSIR
Legal Notice send a blank message with REQUEST LEGAL in the subject line to
Cal...@cs....


This message has been scanned for viruses and dangerous content by MailScanner, 
and is believed to be clean.

Hi,

unfortunately, debs.astraw.com is not accessible.

 wr@[ikarus]:~> ping debs.astraw.com
ping: unknown host debs.astraw.com

 wr@[ikarus]:~> ping www.astraw.com
PING www.astraw.com (66.33.203.253) 56(84) bytes of data.
64 bytes from basic-emu.glass.dreamhost.com (66.33.203.253): icmp_seq=1 ttl=41 
time=169 ms


Am Montag, 23. Oktober 2006 13:18 schrieb Sven Schreiber:
> Willi Richert schrieb:
> > Am Samstag, 21. Oktober 2006 20:30 schrieb Eric Firing:
> >>> matplotlib.__version__ == 0.82
> >>
> >> There is the problem: you need to update your matplotlib.
> >>
> >> Eric
> >
> > Thanks. Unfortunately, I've not yet found a more recent ubuntu/dapper
> > repo for matplotlib. Any experience with installing the ubuntu edgy
> > version (0.87) on dapper?
>
> check out the ubuntu (dapper) section at
> http://new.scipy.org/Installing_SciPy/Linux, that repository (by Andrew
> Straw) also gives you recent matplotlib packages.
> -sven

Willi Richert schrieb:
> Am Samstag, 21. Oktober 2006 20:30 schrieb Eric Firing:
>>> matplotlib.__version__ == 0.82
>> There is the problem: you need to update your matplotlib.
>>
>> Eric
> 
> Thanks. Unfortunately, I've not yet found a more recent ubuntu/dapper repo for 
> matplotlib. Any experience with installing the ubuntu edgy version (0.87) on 
> dapper?
> 

check out the ubuntu (dapper) section at
http://new.scipy.org/Installing_SciPy/Linux, that repository (by Andrew
Straw) also gives you recent matplotlib packages.
-sven

Am Samstag, 21. Oktober 2006 20:30 schrieb Eric Firing:
> >
> > matplotlib.__version__ == 0.82
>
> There is the problem: you need to update your matplotlib.
>
> Eric

Thanks. Unfortunately, I've not yet found a more recent ubuntu/dapper repo for 
matplotlib. Any experience with installing the ubuntu edgy version (0.87) on 
dapper?

Thanks,
wr

Thanks Eric,

That confirms my suspicions. Does anyone know where I can get an rc2
binary for win32? I only see rc3 up on sourceforge.

PS: I think my original e-mail got posted to the list twice... if
that's the case I apologize for the (inadvertent) spam.

On 10/22/06, Eric Firing <ef...@ha...> wrote:
> Todd Page wrote:
> > Hello all,
> >
> > My apologies if this has been brought up and solved before, but I
> > didn't see any solutions that made sense to me in the archives. Here
> > is my stack:
> > python 2.4 (from python.org)
> > numpy-1.0rc3
> > matplotlib-0.87.6
> >
> > Everything seems to install fine. When I attempt to do 'from pylab
> > import *' I get the following error:
> >
> > RuntimeError: module compiled against version 1000002 of C-API but this version
> > of numpy is 1000009
>
> If I remember correctly, mpl 0.87.6 is compiled against numpy-1.0.rc2,
> but rc3 has since come out with a C-API change so that it will not work
> with 0.87.6. So, give numpy-1.0.rc2 a try.  numpy 1.0 is scheduled to
> come out soon, and I expect that shortly thereafter there will be an mpl
> release to match it.  At that point the numpy C API should be genuinely
> stable and it should no longer be necessary to keep future mpl and numpy
> versions in lockstep.
>
> Eric
>


-- 
Todd Page
tw...@gm...

Todd Page wrote:
> Hello all,
> 
> My apologies if this has been brought up and solved before, but I
> didn't see any solutions that made sense to me in the archives. Here
> is my stack:
> python 2.4 (from python.org)
> numpy-1.0rc3
> matplotlib-0.87.6
> 
> Everything seems to install fine. When I attempt to do 'from pylab
> import *' I get the following error:
> 
> RuntimeError: module compiled against version 1000002 of C-API but this version
> of numpy is 1000009

If I remember correctly, mpl 0.87.6 is compiled against numpy-1.0.rc2, 
but rc3 has since come out with a C-API change so that it will not work 
with 0.87.6. So, give numpy-1.0.rc2 a try.  numpy 1.0 is scheduled to 
come out soon, and I expect that shortly thereafter there will be an mpl 
release to match it.  At that point the numpy C API should be genuinely 
stable and it should no longer be necessary to keep future mpl and numpy 
versions in lockstep.

Eric

Hello all,

My apologies if this has been brought up and solved before, but I
didn't see any solutions that made sense to me in the archives. Here
is my stack:
python 2.4 (from python.org)
numpy-1.0rc3
matplotlib-0.87.6

Everything seems to install fine. When I attempt to do 'from pylab
import *' I get the following error:

RuntimeError: module compiled against version 1000002 of C-API but this version
of numpy is 1000009

The import of the numpy version of the nxutils module,
_nsnxutils, failed.  This is is either because numpy was
unavailable when matplotlib was compiled, because a dependency of
_nsnxutils could not be satisfied, or because the build flag for
this module was turned off in setup.py.  If it appears that
_nsnxutils was not built, make sure you have a working copy of
numpy and then re-install matplotlib. Otherwise, the following
traceback gives more details:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "C:\Development\Python25\Lib\site-packages\pylab.py", line 1, in <module>

    from matplotlib.pylab import *
  File "C:\Development\Python25\Lib\site-packages\matplotlib\pylab.py", line 199
, in <module>
    import mlab  #so I can override hist, psd, etc...
  File "C:\Development\Python25\Lib\site-packages\matplotlib\mlab.py", line 64,
in <module>
    import nxutils
  File "C:\Development\Python25\Lib\site-packages\matplotlib\nxutils.py", line 1
7, in <module>
    from matplotlib._ns_nxutils import *
ImportError: numpy.core.multiarray failed to import

I saw an e-mail in the archive that had a similar error, but the
solution was apparently to install scipy-0.51? I don't need or
necessarily want scipy.. is there a reason the latest win32 binaries
for matplotlib and numpy don't play together? Any help is appreciated.
Thanks!

- Todd


-- 
Todd Page
tw...@gm...

using scatter() to draw a plot, is it possible to create a legend for
this?

I have blue points of varying sizes representing values in the plot

I'd like to show a medium-sized blue dot for this plot in the legend.
Using the legend() function shows nothing for scatter plots in the
displayed legend.


Thanks,

Jeff

>>>>> "David" == David Andrews <irb...@gm...> writes:

    David> Any help would be appreciated,

http://www.scipy.org/Cookbook/Matplotlib/Animations

Hi,

I'm slightly confused as to how i'm supposed to write a program that
will call plot() once initially, and then subsequently update the plot
with new points as they are generated.

As i understand it I should be doing something along the following lines:

#First call:
plot(x0,y0,hold=True)
show()

#Subsequent calls:
plot(x1,y1,hold=True)
plot(x2,y2,hold=True)
...


But this does not work as expected.

Any help would be appreciated,

Regards,

Dave Andrews

Hi all,

    am I the only one that noticed that 3D plots do not work anymore? Unless
I am missing something, contour3D, contourf3D simply fails with this
message:

Traceback (most recent call last):
  File "C:\Python25\Lib\site-packages\matplotlib\figure.py", line 440, in
add_ax
es
    a = Axes(self, rect, **kwargs)
  File "C:\Python25\lib\site-packages\matplotlib\axes.py", line 348, in
__init__

    self._position = map(makeValue, rect)
TypeError: argument 2 to map() must support iteration
While the other 3D plots suffer of problems when the mouse enter the plot
window; I get these error messages:

Traceback (most recent call last):
  File "C:\Python25\Lib\site-packages\matplotlib\backends\backend_wx.py",
line 1149, in _onMotion
    FigureCanvasBase.motion_notify_event(self, x, y, guiEvent=evt)
  File "C:\Python25\Lib\site-packages\matplotlib\backend_bases.py", line
895, in motion_notify_event
    func(event)
  File "C:\Python25\lib\site-packages\matplotlib\axes3d.py", line 414, in
on_move
    if event.inaxes != self or not self.M:
ValueError: The truth value of an array with more than one element is
ambiguous.
 Use a.any() or a.all()
Am I missing something? I am using:

Windows XP
Python 2.5
wx* 2.7.1.1 (it doesn't matter in reality)
Numpy 1.0rc2
matplotlib 0.87.6 (NOT svn)

Thanks for every suggestion.


-- 
Andrea.

"Imagination Is The Only Weapon In The War Against Reality."
http://xoomer.virgilio.it/infinity77/

Willi Richert wrote:
> Am Freitag, 20. Oktober 2006 17:31 schrieb John Hunter:
>> from pylab import *
>>
>> ax1 = subplot(111)
>> t = arange(0.01, 10.0, 0.01)
>> s1 = exp(t)
>> plot(t, s1, 'b-')
>> xlabel('time (s)')
>> ylabel('exp')
>>
>>
>> # turn off the 2nd axes rectangle with frameon kwarg
>> ax2 = twinx()
>> s2 = sin(2*pi*t)
>> plot(t, s2, 'r.')
>> ylabel('sin')
>> ax2.yaxis.tick_right()
>> ax2.set_ylim(ymin=-3)
>> show()
> 
> Hi,
> 
> this example does not work for me: Right axis is in the range [-1, 1].
> (The same as the example two_axes.py with an additional set_ylim in my 
> previous mail. Sorry for the sparse information.)
> 
> matplotlib.__version__ == 0.82

There is the problem: you need to update your matplotlib.

Eric

Am Freitag, 20. Oktober 2006 17:31 schrieb John Hunter:
> from pylab import *
>
> ax1 = subplot(111)
> t = arange(0.01, 10.0, 0.01)
> s1 = exp(t)
> plot(t, s1, 'b-')
> xlabel('time (s)')
> ylabel('exp')
>
>
> # turn off the 2nd axes rectangle with frameon kwarg
> ax2 = twinx()
> s2 = sin(2*pi*t)
> plot(t, s2, 'r.')
> ylabel('sin')
> ax2.yaxis.tick_right()
> ax2.set_ylim(ymin=-3)
> show()

Hi,

this example does not work for me: Right axis is in the range [-1, 1].
(The same as the example two_axes.py with an additional set_ylim in my 
previous mail. Sorry for the sparse information.)

matplotlib.__version__ == 0.82
python 2.4

Thanks,
wr

I found that the problem occurs when I have GTKAgg WXAgg or PS backends set in
matplotlibrc file and matplotlib displays fonts properly when using GTK, WX. So
I guess that it must be something wrong with antigrain or preparing text for agg
rendering.

Any help,
regards,
Seweryn Kokot

>>>>> "Jeff" == Jeff Peery <jef...@ya...> writes:

    Jeff> hello, I'm using matplotlib and the wxAgg as a backend with
    Jeff> wxpython for a statistical process control application. I
    Jeff> would like to be able to pick points on an xy line and get
    Jeff> the index of the point. I looked at the picker_demo demo and

You could do something like the following (you'll probably want to add
some epsilon condition for closeness)


from pylab import figure, show, nx

class PickVert:
    def __init__(self, axes, line, callback):
        'call callback with the nearest vert picked'
        self.line = line
        self.axes = axes
        self.axes.figure.canvas.mpl_connect('button_press_event',
    self.pick)
        self.callback = callback

    def pick(self, event):
        if event.button!=1 or event.inaxes!=self.axes: return
        x = line.get_xdata()
        y = line.get_ydata()
        d = nx.sqrt((event.xdata - x)**2 + (event.ydata-y)**2)
        ind = nx.nonzero(d==nx.amin(d))
        self.callback(ind)

def callback(ind):
    print 'the index is', ind

fig = figure()
ax = fig.add_subplot(111)
x = nx.arange(20)
y = nx.mlab.rand(20)
line, = ax.plot(x, y, 'o')

picker = PickVert(ax, line, callback)
show()


    Jeff> I am also a bit confused as to how pylab is related to
    Jeff> matplotlib. I was informed not to use pylab if using a
    Jeff> backend and wxpython application, so I am only using
    Jeff> matplotlib. Is there a similar bind function to use for
    Jeff> binding the pick event to the callback function? thanks.

pylab is a procedural wrapper of the matplotlib API that does things
for you like manage the creation and destruction of GUI windows and it
works across multiple GUIs.  If you are writing a GUI app, you do not
want pylab to manage that for you.

JDH

Hello,

I have the following script and in matplotlib 0.87.5 (debian sid), text is not
correctly displayed, whereas in matplotlib 0.82-1 it was ok. See attached
pictures.

#!/usr/bin/python
# -*- coding: iso-8859-2 -*-
from pylab import *

ab = unicode('żłóńę','iso-8859-2')

plot([1, 3, 4],[1, 4, 2])
xlabel(ab)
show()

Any idead what is wrong
Regards,
Seweryn Kokot

I don't know how to make the "//" symbol in the y-axis, but if you  
have two plots that share the same x-axis, you can represent this  
kind of data.

The "Working with multiple figure and axes" section of the tutorial

http://matplotlib.sourceforge.net/tutorial.html

is almost right; if you turn off the display of the x-axis in the  
upper plot, it will be even better.

&C

On Oct 19, 2006, at 3:05 PM, CL wrote:

> Hi, group,
>              I am wondering if matplotlib can make a plot look like  
> the attached one. Basically, the y-axis was collapsed to show both  
> one extreme value and the rest of smaller values.
>
> Thanks.
>
> Chunlei
>
> <moz-screenshot.jpg>
>
> <moz-screenshot.jpg>
> ---------------------------------------------------------------------- 
> ---
> Using Tomcat but need to do more? Need to support web services,  
> security?
> Get stuff done quickly with pre-integrated technology to make your  
> job easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache  
> Geronimo
> http://sel.as-us.falkag.net/sel? 
> cmd=lnk&kid=120709&bid=263057&dat=121642______________________________ 
> _________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

hello, I'm using matplotlib and the wxAgg as a backend with wxpython for a statistical process control application. I would like to be able to pick points on an xy line and get the index of the point. I looked at the picker_demo demo and I have an idea of how this works but I'm a bit confused as to how to pick a vertex on a line, or if that isn't possible how to pick a marker created by scatter(). I am also a bit confused as to how pylab is related to matplotlib. I was informed not to use pylab if using a backend and wxpython application, so I am only using matplotlib. Is there a similar bind function to use for binding the pick event to the callback function? thanks.

Jeff

 		
---------------------------------
Get your email and more, right on the  new Yahoo.com 

>>>>> "Khem" == Khem Raj <raj...@gm...> writes:

    Khem> Hi I am new to matplotlib and I am trying to use it to draw
    Khem> a matrix. I would want to give a different color to each box
    Khem> depending upon the value this matrix element has


    Khem> Could someone help me in quick start meanwhile I am looking
    Khem> into documentation.

figure()
imshow(rand(10,5), cmap=cm.jet, interpolation='nearest', aspect='auto')
colorbar()

figure()
pcolor(rand(10,5), cmap=cm.hot)


JDH

>>>>> "Johannes" == Johannes Elbs <ml....@gm...> writes:

    Johannes> Hi, I have time data stored in the format hh:mm:ss and
    Johannes> would like to plot it, having about 4 to 6 ticks labeled
    Johannes> in this format.  What I'm doing at the moment is to
    Johannes> first convert my time to seconds and plot y as function
    Johannes> of yime in seconds.

You first need to convert them to matplotlib dates.  Assume your data
are collected on Oct 10th, 2006

  import matplotlib.numerix as nx
  import datetime
  from matplotlib.dates import date2num, MINUTES_PER_DAY, SEC_PER_DAY

  d0 = date2num(datetime.date(2006,10,10))
  def convert(s):
      h,m,s = map(float, s.split(':'))
      return d0 + h/24. + m/MINUTES_PER_DAY + s/SEC_PER_DAY


In [11]: convert('12:32:02')
Out[11]: 732594.52224537032

matplotlib represents dates as days (float) since 0000-00-00

Now you need to convert your list of strings to mpl dates.  If your
list of strings is called datestrs

  dates = nx.array([convert(s) for s in datestrs])

Now when you plot you need to tell mpl you are plotting dates

  ax.plot_date(dates, vals)

We have pretty smart tick locators and tick formatters for date
plotting, so you may be happy with what you get, but if not you can
set your own

  from matplotlib.dates import DateFormatter
  ax.xaxis.set_major_formatter('%H:%M:%S')

I also find it helpful on date plots to do the following

   # move the subplot up a bit to make for rotated labels
   fig.subplots_adjust(bottom=0.15)
   for label in ax.get_xticklabels():
       label.set_rotation(30)
       label.set_horizontalalignment('right')
  

see also

  http://matplotlib.sf.net/matplotlib.dates.html
  http://matplotlib.sf.net/matplotlib.ticker.html

the users guide and the following examples

date_demo1.py
date_demo2.py
date_demo_convert.py
date_demo_rrule.py
finance_demo.py


JDH

>>>>> "Willi" == Willi Richert <w.r...@gm...> writes:

    Willi> Hi, I am plotting two graphs in one according to
    Willi> http://matplotlib.sourceforge.net/examples/two_scales.py

    Willi> I want the plot belonging to the right axis to start with
    Willi> y=0. However,

    Willi> ax2.set_ylim(ymin=0.0)

    Willi> does not affect the plot at all.

It always helps to post a complete example; otherwise we have to use
our amazing powers of deduction to figure out what you might be doing
wrong.  In this case, two possibilities come to mind:

  1) you are calling set_ylim before a plot command and the autoscaler
     is kicking on the plot command and overriding your changes.  You
     should call set_ylim after all plot commands, or turn autoscaling
     off with the autoscale_on property of the Axes

  2) you are working in interactive mode, eg ipython -pylab, and the
     draw command is not being triggered since you are making an OO
     call.  You can force a draw command in pylab with "draw" or with
     fig.canvas.draw().

The following *does* work for me, so maybe you can follow it as an
example

from pylab import *

ax1 = subplot(111)
t = arange(0.01, 10.0, 0.01)
s1 = exp(t)
plot(t, s1, 'b-')
xlabel('time (s)')
ylabel('exp')


# turn off the 2nd axes rectangle with frameon kwarg
ax2 = twinx()
s2 = sin(2*pi*t)
plot(t, s2, 'r.')
ylabel('sin')
ax2.yaxis.tick_right()
ax2.set_ylim(ymin=-3)
show()

I am sure this (superimposed Line and Bar plots) is
easy to do, but I cannot see any discussion of "how"
on the user list (difficult to know which phrase to search!)
 
Any pointers?
 
Thanks
Derek

-- 
This message is subject to the CSIR's copyright, terms and conditions and
e-mail legal notice. Views expressed herein do not necessarily represent the
views of the CSIR.
 
CSIR E-mail Legal Notice
http://mail.csir.co.za/CSIR_eMail_Legal_Notice.html 
 
CSIR Copyright, Terms and Conditions
http://mail.csir.co.za/CSIR_Copyright.html 
 
For electronic copies of the CSIR Copyright, Terms and Conditions and the CSIR
Legal Notice send a blank message with REQUEST LEGAL in the subject line to
Cal...@cs....


This message has been scanned for viruses and dangerous content by MailScanner, 
and is believed to be clean.

I have a simple, but frustrating problem.
 
I need to plot a number of series onto a boxplot.  Some sample
code looks as follows:
 
from pylab import *
from scipy import array
jan = array([46, 38, 28, 14, 11, 7,  4,  3, 3, 3, 3, 3, 3, 2, 2, 1])
#jan = array([46, 38, 28, 14, 11, 7,  4,  3, 3, 3, 3, 3]) #this length
works!!
feb = array([ 2, 12,  3,  3, 13, 10, 17, 6, 6, 6, 3, 3])
jan.shape = (-1, 1)
feb.shape = (-1, 1)
series = concatenate( (jan, feb), 1 )
figure()
boxplot(series,1)
show()
 
(In Real Life, of course, the data arrays are built up elsewhere
and passed into the plotting function...)
 
The problem is that I get the error:
ValueError: array dimensions must agree except for d_0
 
because jan and feb have different lengths - however, I have no
control
over the size of these arrays - they vary greatly according to
external
issues.
 
How do I plot multiple. variable-length array on the same boxplot?
 
Thanks
Derek

-- 
This message is subject to the CSIR's copyright, terms and conditions and
e-mail legal notice. Views expressed herein do not necessarily represent the
views of the CSIR.
 
CSIR E-mail Legal Notice
http://mail.csir.co.za/CSIR_eMail_Legal_Notice.html 
 
CSIR Copyright, Terms and Conditions
http://mail.csir.co.za/CSIR_Copyright.html 
 
For electronic copies of the CSIR Copyright, Terms and Conditions and the CSIR
Legal Notice send a blank message with REQUEST LEGAL in the subject line to
Cal...@cs....


This message has been scanned for viruses and dangerous content by MailScanner, 
and is believed to be clean.

Hi,

I have time data stored in the format hh:mm:ss and would like to plot it,
having about 4 to 6 ticks labeled in this format.
What I'm doing at the moment is to first convert my time to seconds and plot
y as function of yime in seconds.

Here the first question: Is this the way to go or is there a function I
missed which allows to directly plot from the former format?

Now I'd like to have the labels to be displayed in the hh:mm:ss format, in a
way that it automatically adopts when zooming/dezooming. Without zooming I
figured out that it is quite easy to use the xticks method, but is there a
possibility to write a function which automatically readadopts the tick
spacing?
If not, I found in the manual the dateformatter class, but if I understand
correctly, I first have to transfer the time to strftime? Or can I directly
use this class using the time starting from midnight?

Thanks for any response,
Johannes