matplotlib-users Mailing List for matplotlib

clean completely your precedent installation. That means remove the directo=
ry=20
matplotlib in the site-package directory.

N.

Le mardi 10 octobre 2006 09:46, Bill Dandreta a =E9crit=A0:
> I upgraded matplotlib (from 0.84 to 0.87.4 and then 0.87.5) and my
> scripts that previously worked no longer plot.
>
> No error message just won't plot.
>
> I tried down grading to 0.84, that no longer plots either.
>
> I used the simple_plot.py example for testing.
>
> Bill
>
>
>
> -------------------------------------------------------------------------
> 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=3Djoin.php&p=3Dsourceforge&CID=3D=
DEVDEV
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Try deleting you "~/.matplotlib" folder.  Also, the latest matplotlib
is 0.87.6 and it should be used with numpy-1.0rc1 and up.  The older
builds require older versions of numpy.  (This is just assuming you
are using numpy, but an important note for all).

On 10/10/06, Bill Dandreta <wjd...@at...> wrote:
> I upgraded matplotlib (from 0.84 to 0.87.4 and then 0.87.5) and my
> scripts that previously worked no longer plot.
>
> No error message just won't plot.
>
> I tried down grading to 0.84, that no longer plots either.
>
> I used the simple_plot.py example for testing.
>
> Bill
>
>
>
> -------------------------------------------------------------------------
> 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
>

On 10/10/06, R. Padraic Springuel <R.S...@um...> wrote:
> I haven't tried the binaries because there isn't one for python 2.5 and
> Mac OS 10.4 (the particular combination that I have).  I could use the
> binaries if I went back to python 2.4, but I'm trying avoid that if
> possible.

So actually there is a build for you... actually two.  I built
"matplotlib-0.87.6-py2.5-macosx-10.3-fat.egg" with python2.5 on OS
10.4.  For some strange reason setuptools or distutils tags it with
10.3.  It might be because the build targets to 10.3.9 SDK, so it
should be compatible with 10.3.9 and up.
"matplotlib-0.87.6-py2.5-macosx10.4.zip" is the exact same build but
bundled with bdist_mpkg.  It is understanding how this is confusing.
I am cc'ing the pythonmac-sig list to see if someone can better
explain the thought process behind this.

- Charlie

I upgraded matplotlib (from 0.84 to 0.87.4 and then 0.87.5) and my
scripts that previously worked no longer plot.

No error message just won't plot.

I tried down grading to 0.84, that no longer plots either.

I used the simple_plot.py example for testing.

Bill

I haven't tried the binaries because there isn't one for python 2.5 and 
Mac OS 10.4 (the particular combination that I have).  I could use the 
binaries if I went back to python 2.4, but I'm trying avoid that if 
possible.
-- 

R. Padraic Springuel
Teaching Assistant
Department of Physics and Astronomy
University of Maine
Bennett 309
Office Hours: Wednesday 2-3pm

>>>>> "humufr" == humufr  <hu...@ya...> writes:

    humufr> 		Hi, I'm using very often the load command, I
    humufr> would like to know why the default comments character is
    humufr> still '%', the one from matlab, other than historical?  I
    humufr> would prefer to have the one from python '#'. I know it's
    humufr> just a personal preference but we are using python so for
    humufr> consistency I'm thinking that it would be better to have
    humufr> '#' or at least to have an option in matplotlibrc to
    humufr> choose the default character.

Good call -- changed in svn and duly noted in the CHANGELOG and in
API_CHANGES

JDH

		Hi,

I'm using very often the load command, I would like to know why the default 
comments character is still '%', the one from matlab, other than historical? 
I would prefer to have the one from python '#'. I know it's just a personal 
preference but we are using python so for consistency I'm thinking that it 
would be better to have '#' or at least to have an option in matplotlibrc to 
choose the default character.

Thanks,

N. 

Hi Eric,

I don't think this is a bug. Your last import doesnt do what you think it is 
doing. For example: 

In [1]: import matplotlib as d

In [2]: from d import rcParams
---------------------------------------------------------------------------
exceptions.ImportError                               Traceback (most recent 
call last)

/home/darren/<ipython console>

ImportError: No module named d


You have imported the module to namespace d, you have not renamed the module. 
If you want median and std in the main namespace, you could do

median = num.median
std = num.std

that is, you could this if median and std were provided by numerix, which they 
are not. They are in matplotlib.numerix.mlab.

Darren


On Tuesday 10 October 2006 08:48, Eric Emsellem wrote:
> Hi,
>
> I have a problem when importing numpy from matplotlib.numerix:
>
>              When I do:
>
> import matplotlib
> import matplotlib.numerix as num # numpy
> from num import median, std
>
>               I get:
> exceptions.ImportError                               Traceback (most
> recent call last)
> ImportError: No module named num
>
> It seems that it does not import numpy but something else. (for example
> num.std does not exist).
> How do I solve that? (I saw a thread which seems to indicate that it was
> solved recently, but I have recent versions of everything :
>     matplotlib version 0.87.5
>     platform is linux2
>     numerix numpy 1.0rc1.dev3157
>     backend GTKAgg version 2.8.2
>     Python 2.4.2 (#1, May  2 2006, 08:13:46)
>     IPython 0.7.3.svn -- An enhanced Interactive Python.
>
> What could I do there?
>
> Note that in the help of matplotlib.numerix, numpy is only indicated in
> Option 0. (see below)
>
> thanks for any help there.
>
> Eric
> ==================================================
> when I do:
>
> help num
>
> ==> I get:
>
> Help on package matplotlib.numerix in matplotlib:
>
> NAME
>     matplotlib.numerix - numerix  imports either Numeric or numarray
> based on various selectors.
>
> FILE
>     /usr/local/lib/python2.4/site/packages/matplotlib/numerix/__init__.py
>
> DESCRIPTION
>     0.  If the value "--numpy","--numarray" or "--Numeric" is specified
> on the
>         command line, then numerix imports the specified
>         array package.
>
>     1. The value of numerix in matplotlibrc: either Numeric or numarray
>
>     2. If none of the above is done, the default array package is Numeric.
>        Because the matplotlibrc always provides *some* value for numerix
>        (it has it's own system of default values), this default is most
>        likely never used.
>
>     To summarize: the  commandline is examined first, the  rc file second,
>     and the default array package is Numeric.
>
>
> -------------------------------------------------------------------------
> 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

-- 
Darren S. Dale, Ph.D.
Staff Scientist
Cornell High Energy Synchrotron Source
Cornell University
275 Wilson Lab
Rt. 366 & Pine Tree Road
Ithaca, NY 14853

dd...@co...
office: (607) 255-3819
fax: (607) 255-9001
http://www.chess.cornell.edu

Hi,

I have a problem when importing numpy from matplotlib.numerix:

             When I do:

import matplotlib
import matplotlib.numerix as num # numpy
from num import median, std

              I get:
exceptions.ImportError                               Traceback (most
recent call last)
ImportError: No module named num

It seems that it does not import numpy but something else. (for example
num.std does not exist).
How do I solve that? (I saw a thread which seems to indicate that it was
solved recently, but I have recent versions of everything :
    matplotlib version 0.87.5
    platform is linux2
    numerix numpy 1.0rc1.dev3157
    backend GTKAgg version 2.8.2
    Python 2.4.2 (#1, May  2 2006, 08:13:46)
    IPython 0.7.3.svn -- An enhanced Interactive Python.

What could I do there?

Note that in the help of matplotlib.numerix, numpy is only indicated in
Option 0. (see below)

thanks for any help there.

Eric
==================================================
when I do:

help num

==> I get:

Help on package matplotlib.numerix in matplotlib:

NAME
    matplotlib.numerix - numerix  imports either Numeric or numarray
based on various selectors.

FILE
    /usr/local/lib/python2.4/site/packages/matplotlib/numerix/__init__.py

DESCRIPTION
    0.  If the value "--numpy","--numarray" or "--Numeric" is specified
on the
        command line, then numerix imports the specified
        array package.

    1. The value of numerix in matplotlibrc: either Numeric or numarray

    2. If none of the above is done, the default array package is Numeric.
       Because the matplotlibrc always provides *some* value for numerix
       (it has it's own system of default values), this default is most
       likely never used.

    To summarize: the  commandline is examined first, the  rc file second,
    and the default array package is Numeric.

Le Lundi 9 Octobre 2006 19:11, Zack a =E9crit=A0:=0A=
> Bonjour Fran=E7ois!=0A=
>=0A=
> Take it easy.=0A=
>=0A=
> rc("lines", markeredgewidth=3D4)=0A=
>=0A=
> There is no reasons to put cycle inside you code.=0A=
>=0A=
> Regards,=0A=
=0A=
Thanks for the tips zack !=0A=
It works like a dream ...=0A=
=0A=
Is there a way to specify only major or minor ticks marks ?=0A=
=0A=
--=20=0A=
Fran=E7ois BEAUBERT=0A=
ENSIAME/LME=0A=
Bureau A9=0A=
Universit=E9 de Valenciennes - Le Mont Houy - 59313 Valenciennes Cedex 9=0A=
T=E9l=E9phone: (0/33)3 27 51 19 72=0A=
T=E9l=E9copie: (0/33)3 27 51 19 61=0A=
Mel: fra...@un...=0A=
=0A=
--=20=0A=
Ce message a =E9t=E9 v=E9rifi=E9 par MailScanner=0A=
pour des virus ou des polluriels et rien de=0A=
suspect n'a =E9t=E9 trouv=E9.=0A=
MailScanner remercie transtec pour son soutien.=0A=
=0A=

i need a tutorial on plotting countour lines=0Agave me a url=0A=0A=0A=0A=0A=
=09=09=0A__________________________________________________________=0AYahoo=
! India Answers: Share what you know. Learn something new=0Ahttp://in.answe=
rs.yahoo.com/

> data (and I assume your graphs are done a similar
> way), and so I 

It turned out to be too slow for me to generate them
on the fly. I pregenerate most images that you see at
www.eswap.com

If you have a fast enought web connection the page
should load up in 3-4 secnods. 

Vineet


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

On 10/9/06, zh...@um... <zh...@um...> wrote:
> I mean in the original data format, without any characters like "[],"
> Thanks
>
>
>
> -------------------------------------------------------------------------
> 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
>

Does this help:

a = [[1, 2, 3], [4, 5, 6]]
f = open("output.txt", "w")
for row in a:
    f.write(" ".join(row))
    f.write("\n")
f.close()

Check this also:
http://effbot.org/librarybook/csv.htm

HTH,
Edin

hu...@ya... wrote:
> Hi,
> 
> some times ago I saw that the rgb color was not working anymore, the hexa code 
> yes but not the rgb.
> 
> markerfacecolor=(120,120,120,)
> 
> retrun this error:
> 
> ValueError: to_rgb: Invalid rgb arg "(120, 120, 120)"
> to_rgb: Invalid rgb arg "(120, 120, 120)"

This really is an invalid rgb arg; a valid equivalent would be 
(120.0/255, 120.0/255, 120.0/255).  Matplotlib colors can be specified 
in several ways, including triplets of numbers from 0 to 1, but not 
including triplets of numbers from 0 to 255.

Eric

> what command is used to "pull" the graph into the
> web page - the matplotlib "show" command?

There are a bunch of ways to do this. One way is to pregenerate the 
charts you need and store them as .png or .gif files on the server. If 
you need truly dynamic images, then you need to make a Python program 
(WSGI, CGI, whatever) that returns the image/png on the fly. It's not 
easy to do this with matplotlib, but it's not too hard; search the 
mailing list archives for notes about getting the image bits in memory 
rather than drawing on screen or writing to disk.

> PS <1 sec to make a chart sounds quick - given that it takes much
> longer than  1 second to load the chart across the web into one's browser...
>
>   

All depends on your load and expectations. If the 1 second is all CPU 
time, then 1 second means you can only generate 100,000 or so graphs a 
day per server.

"Derek Hohls" <DH...@cs...> writes:

> ie. 0.2  0.4   0.6... followed by a rather large and ugly-looking
> x1e+4
> Is there is a simple way to avoid exponential notation?

Try using a FormatStrFormatter with a suitable format string:

  from matplotlib.ticker import FormatStrFormatter
  ax=gca()
  ax.xaxis.set_major_formatter(FormatStrFormatter('%5.0f'))

For more information, see
http://matplotlib.sourceforge.net/matplotlib.ticker.html
and grep for Formatter in the examples directory.

-- 
Jouni

I do not see a reason why the problem should be related. If the problem
persists with the latest SVN version, could you please file a bug report?=




hu...@ya... wrote:
> Hi,
>
> some times ago I saw that the rgb color was not working anymore, the he=
xa code=20
> yes but not the rgb.
>
> markerfacecolor=3D(120,120,120,)
>
> retrun this error:
>
> ValueError: to_rgb: Invalid rgb arg "(120, 120, 120)"
> to_rgb: Invalid rgb arg "(120, 120, 120)"
>
> but:
>
> markerfacecolor=3D'#888888'
>
> is working fine
>
> I send a mail, I don't remember if I had time to post a bug report. Per=
haps is=20
> it link?
>
> N.
>
> Le dimanche 8 octobre 2006 15:49, Eric Firing a =E9crit :
>  =20
>> Stefan,
>>
>> Thanks for the bug report--it is surprising that no one noticed this
>> earlier.  It looks like Norbert's changes completely wiped out the
>> setting of markerfacecolor and markeredgecolor.  I have reverted axes.=
py
>> in svn to the 2788 revision (via revision 2799) so that it will work
>> until Norbert gets a chance to add marker handling back into his patch=

>> and reapply it.
>>
>> Eric
>>
>> Stefan van der Walt wrote:
>>    =20
>>> Hi all
>>>
>>> The marker behaviour changed in
>>>
>>> ---------------------------------------------------------------------=
---
>>> r2790 | nnemec | 2006-09-29 11:46:57 +0200 (Fri, 29 Sep 2006) | 1 lin=
e
>>>
>>> reworked linestyle and markercolor handling
>>>
>>>
>>> For example, try the following:
>>>
>>> N.plot(N.random.random(1000),'r.')
>>>
>>> Note that all the dots are blue!
>>>      =20
>> ----------------------------------------------------------------------=
---
>> Take Surveys. Earn Cash. Influence the Future of IT
>> Join SourceForge.net's Techsay panel and you'll get the chance to shar=
e
>> your opinions on IT & business topics through brief surveys -- and ear=
n
>> cash
>> http://www.techsay.com/default.php?page=3Djoin.php&p=3Dsourceforge&CID=
=3DDEVDEV
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>    =20
>
> -----------------------------------------------------------------------=
--
> 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=3Djoin.php&p=3Dsourceforge&CID=3D=
DEVDEV
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>
>  =20

Hi Stefan,

as of rev. 2800, the problem should be solved. For more details, see the
discussion about "marker color handling" on matplotlib-devel.

Greetings,
Norbert


Stefan van der Walt wrote:
> Hi all
>
> The marker behaviour changed in
>
> -----------------------------------------------------------------------=
-
> r2790 | nnemec | 2006-09-29 11:46:57 +0200 (Fri, 29 Sep 2006) | 1 line
>
> reworked linestyle and markercolor handling
>
>
> For example, try the following:
>
> N.plot(N.random.random(1000),'r.')
>
> Note that all the dots are blue!
>
> Regards
> St=E9fan
>
> -----------------------------------------------------------------------=
--
> 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=3Djoin.php&p=3Dsourceforge&CID=3D=
DEVDEV
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>
>  =20

Vineet

Sorry - I was not too clear on my question; I *am* currently able
to use matplotlib - what I am not sure how to is to integrate 
it into a web framework.  I have been looking at using Pylons - 
as that is WSGI compliant - but probably any one will do.
 
I would like to be able to generate graphs "on the fly" from database 
data (and I assume your graphs are done a similar way), and so I 
was wondering what are the specifics of the approach you used
for this - is it just a case of installing Python onto a server,
plus the web framework code, and then importing the matplotlib
code - if so, what command is used to "pull" the graph into the
web page - the matplotlib "show" command? - or do you store the
graph somewhere on a temporary basis and then delete again when
done??
 
Thanks
Derek

PS <1 sec to make a chart sounds quick - given that it takes much
longer 
than  1 second to load the chart across the web into one's browser...

>>> Vinj Vinj <vin...@ya...> 2006/10/09 04:25 PM >>>

> Any change you could share back with the community
the approach
> you followed - web framework used?  integration
strategy?  deploy-
> ment and setup issues?  graph creation methodology?

web framework used: cherrypy
deploymnet os: ubuntu
setup issues: none

> I am looking to tackle a project with multi-chart
reporting 
> component features and would really love to use
Matplotlib
> with it.... but am not quite sure where to start. 

matplotlib comes with a lot of examples and that's
probably the best
place to start. There is an example for how to  save
charts to a file
(which is what I do) and another one for streaming the
charts back to
the user.

The only issue I have faced is that matplotlib is slow
for my chart types (takes about 0.8 seconds to do a
chart). I'm hoping that future version of matplotlib
will be faster. 

Guys at stockcharts.com generate charts in about 0.3
seconds. So matplotlib (for the charts I generate) is
about 2.5 slower than what they use. For most people
that might not be an issue, but since I pre-generate
100,000s of graphs daily it can cause a lot of pain. 

Vineet


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


-- 
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.