matplotlib-users Mailing List for matplotlib

Thank you all for your help.
Regards,
Sam
Matthias Michler wrote:
> 
> Hi Sam,
> 
> On Monday 28 April 2008 05:09:36 samwo wrote:
>> Hi.
>>
>> I run these codes in ipython command line interface.
>>
>> from pylab import *
>> subplot(211)
>> plot([1,2,3],[1,2,3])
> There is a small typo (or a wrong method):
>> set(gca(), xtickslabel = [])
> needs to be 
> setp(gca(), xticklabels = [])
> 
>> and got the error saying "Type error: set() does not take keyword
>> arguments" My OS is windows XP. The code I used is copied from matplotlib
>> tutorial. Pls help explain why it didnt work.
> 
> regards Matthias
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
> Don't miss this year's exciting event. There's still time to save 100ドル. 
> Use priority code J8TL2D2. 
> http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
> 
-- 
View this message in context: http://www.nabble.com/pls-help%3A-error-using-set%28gca%28%29%2C...%29-tp16930305p16989834.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

Hi,
I would like to generate (or simulate) 3D histogram... How I can do this with matplotlib/pylab (if I can) ?
I tried different ways :
 - 2D histogram in 3D but the result is not usable for me (because I have lot of small bars)
 - generation of bars with, for each bar, 5 Polygons but I didn't success (I don't hunderstand the main principle of the 3D module) and it seems to be slow (I must draw approximatively 300 bars)
 - simulation of bars with large lines (like in Gnuplot, see below) but, again, I don't hunderstand the main principle of the 3D module and I don't know how I can plot set of lines in a 3D environment...
Today, I do this with Gnuplot :
>>> splot "hist3D.dat" with boxes linewidth 10
hist.dat :
0 0 3
0 1 2
0 2 5
1 0 1
1 1 4
1 2 1
2 0 5
2 1 2
2 2 3
Does someone have an idea ?
Thanks

Hello,
I have decided to switch to the QtAgg backend because the Designer
software is very appealing. I succeeded in adding a custom widget
representing the FigureCanvasQTAgg, but when I compile the ui file, I
see that it tries to instantiate it as FigureCanvasQTAgg(Parent) where
Parent is the parent widget, which seems to be the standard format for
instantiating widgets. However, FigureCanvasQTAgg expects a figure as
an argument.
I have seen this tutorial:
http://www.scipy.org/Cookbook/Matplotlib/Qt_with_IPython_and_Designer
which suggests making a MatplotlibWidget to wrap the Canvas and set
everything up including the parent. I have written a version of this
for Qt4, but this seems like something that should be included in the
Qt4 backends. Any plans on adding this? I'd be happy to contribute.
Attached is my initial version.
Glenn

hello,
so the freetype.rc is defined with env variables that are supposed to be 
fed in when packaging it. LIBZ seemingly was missing when the fedora 8 
packaging of freetype was done. I added -lz directly to freetype.rc and 
rebuilt successfully matplotlib. I have no clue which mailing list to 
ping for a fedora packaging issue....
Anyway, end of story for this part of the thread....
best,
Johann
John Hunter wrote:
> On Tue, Apr 29, 2008 at 2:06 AM, Johann Cohen-Tanugi
> <co...@sl...> wrote:
> 
>> hi John,
>> for the record, I have a fedora 8 box, and the shipped version of
>> pkgconfig (pkgconfig-0.22-4.fc8) gives me :
>>
>> [cohen@jarrett matplotlib]$ /usr/bin/pkg-config --libs freetype2
>> -lfreetype
>>
>> No -lz here, if I am not wrong in my querying this tool .....
>> 
>
> This looks like the correct incantation, so it appears that your
> proplem is a misconfigured freetype2.pc. You might want to post
> something to the fedora list, or to the free type list, or simply make
> a patch against the pc file. I am not sure who maintains that part.
>
> JDH
> 

cheers,
Johann

Hello
I think that my problem is linked to the data struture that I'm using for
the dates since my error message is : "ValueError: setting an array element
with a sequence."
I use list data structure, the one that you can declare through :
datetime_list = [].
And maybe I souldn't use this type of data structure and another one.
Regards
Johan Mazel
*
*2008年4月29日 Johan Mazel <joh...@gm...>:
> Ok.
> I tried to do the job through the old way and it's working.
> But I have the latest version (0.91.2). Is it possible that my version
> wasn't successfully installed and that an older one is doing the job ?
> Anyway, thanks a lot for the (fast) help.
> Johan
>
> 2008年4月29日 John Hunter <jd...@gm...>:
>
> > On Tue, Apr 29, 2008 at 12:10 PM, Johan Mazel <joh...@gm...>
> > wrote:
> > > When you wrote "import matplotlib.ticker", you meant "from
> > matplotlib.dates
> > > import ticker" ?
> >
> > No sorry, I meant "import matplotlib.ticker as ticker"
> >
> > > I think I have a bug or something because I get an error like
> > "ValueError:
> > > setting an array element with a sequence." with a lot of debug stuff
> > from
> > > python and MatPlotLib
> > > The line concerned is the one where I do
> > plot(datetime_list,data_list).
> > > It's very strange because just before I use "my_datetime.strftime("%d
> > %H %M
> > > %S")" with my_datetime as an element of my list and it works
> > perfectly.
> >
> >
> > Possibly your matplotlib version is a bit old. With the latest
> > release (0.91.2) you can pass sequences of native datetimes in. For
> > older versions, you had to convert your dates to numbers first:
> >
> > import matplotlib.dates as mpldates
> >
> > d = mpldates.date2num(datetime_list)
> > ax.plot_date(d, ydata)
> >
> > JDH
> >
>
>

Ok.
I tried to do the job through the old way and it's working.
But I have the latest version (0.91.2). Is it possible that my version
wasn't successfully installed and that an older one is doing the job ?
Anyway, thanks a lot for the (fast) help.
Johan
2008年4月29日 John Hunter <jd...@gm...>:
> On Tue, Apr 29, 2008 at 12:10 PM, Johan Mazel <joh...@gm...>
> wrote:
> > When you wrote "import matplotlib.ticker", you meant "from
> matplotlib.dates
> > import ticker" ?
>
> No sorry, I meant "import matplotlib.ticker as ticker"
>
> > I think I have a bug or something because I get an error like
> "ValueError:
> > setting an array element with a sequence." with a lot of debug stuff
> from
> > python and MatPlotLib
> > The line concerned is the one where I do plot(datetime_list,data_list).
> > It's very strange because just before I use "my_datetime.strftime("%d %H
> %M
> > %S")" with my_datetime as an element of my list and it works perfectly.
>
>
> Possibly your matplotlib version is a bit old. With the latest
> release (0.91.2) you can pass sequences of native datetimes in. For
> older versions, you had to convert your dates to numbers first:
>
> import matplotlib.dates as mpldates
>
> d = mpldates.date2num(datetime_list)
> ax.plot_date(d, ydata)
>
> JDH
>

---------- Forwarded message ----------
From: Johan Mazel <joh...@gm...>
Date: 2008年4月29日
Subject: Re: [Matplotlib-users] Plot data with specified time (hour, minute,
second)
To: John Hunter <jd...@gm...>
When you wrote "import matplotlib.ticker", you meant "from matplotlib.dates
import ticker" ?
I think I have a bug or something because I get an error like "ValueError:
setting an array element with a sequence." with a lot of debug stuff from
python and MatPlotLib
The line concerned is the one where I do plot(datetime_list,data_list).
It's very strange because just before I use "my_datetime.strftime("%d %H %M
%S")" with my_datetime as an element of my list and it works perfectly.
Anyway, thanks a lot for the help.
Johan Mazel
2008年4月29日 John Hunter <jd...@gm...>:
> On Tue, Apr 29, 2008 at 11:38 AM, Johan Mazel <joh...@gm...>
> wrote:
> > Ok.
> > I manage to get either time or date but not both inside the same object
> .
> > And I don't know how to use plot with two object (one for the time and
> the
> > other for the date).
>
> you can get the date and time into a datetime object as follows:
>
> import datetime
> dt = datetime.datetime(2008,4,22,3,45,21)
> print dt.strftime('%Y-%m-%d %H:%M:%S')
>

On Tue, Apr 29, 2008 at 11:38 AM, Johan Mazel <joh...@gm...> wrote:
> Ok.
> I manage to get either time or date but not both inside the same object .
> And I don't know how to use plot with two object (one for the time and the
> other for the date).
you can get the date and time into a datetime object as follows:
import datetime
dt = datetime.datetime(2008,4,22,3,45,21)
print dt.strftime('%Y-%m-%d %H:%M:%S')

Ok.
I manage to get either time or date but not both inside the same object .
And I don't know how to use plot with two object (one for the time and the
other for the date).
Thanks for the answer.
Johan Mazel
2008年4月29日 John Hunter <jd...@gm...>:
> On Tue, Apr 29, 2008 at 8:12 AM, Johan Mazel <joh...@gm...>
> wrote:
>
> > I also would like to know wether I can use the same type of
> functionnality
> > as in Gnuplot where you can specify the format of data in input and the
> data
> > displayed of the x axis. I just guess that the use of the date object
> > replace any kind of setup in data input but anyway...
>
> If you are passing in datetime objects for plotting on the x-axis, you
> can format the tick labels any way you want:
>
> import matplotlib.ticker
>
> formatter = ticker.DateFormatter('%Y-%m-%d %H:%M:%S')
> ax.xaxis.set_major_formatter(formatter)
>
> See the date_demo*.py examples at http://matplotlib.sf.net/examples
>

On Sun, Apr 27, 2008 at 10:09 PM, samwo <su...@gm...> wrote:
>
> Hi.
>
> I run these codes in ipython command line interface.
>
> from pylab import *
> subplot(211)
> plot([1,2,3],[1,2,3])
> set(gca(), xtickslabel = [])
"set" is a python built-in (the mathematical set) and matplotlib now
uses "setp" (for set property) instead.
JDH

Hi Sam,
On Monday 28 April 2008 05:09:36 samwo wrote:
> Hi.
>
> I run these codes in ipython command line interface.
>
> from pylab import *
> subplot(211)
> plot([1,2,3],[1,2,3])
There is a small typo (or a wrong method):
> set(gca(), xtickslabel = [])
needs to be 
setp(gca(), xticklabels = [])
> and got the error saying "Type error: set() does not take keyword
> arguments" My OS is windows XP. The code I used is copied from matplotlib
> tutorial. Pls help explain why it didnt work.
regards Matthias

On Tue, Apr 29, 2008 at 8:12 AM, Johan Mazel <joh...@gm...> wrote:
> I also would like to know wether I can use the same type of functionnality
> as in Gnuplot where you can specify the format of data in input and the data
> displayed of the x axis. I just guess that the use of the date object
> replace any kind of setup in data input but anyway...
If you are passing in datetime objects for plotting on the x-axis, you
can format the tick labels any way you want:
import matplotlib.ticker
formatter = ticker.DateFormatter('%Y-%m-%d %H:%M:%S')
ax.xaxis.set_major_formatter(formatter)
See the date_demo*.py examples at http://matplotlib.sf.net/examples

On Tue, Apr 29, 2008 at 7:45 AM, Michael Droettboom <md...@st...> wrote:
> I was able to reproduce it on 0.90.1 and 0.91.2, but not SVN trunk. This
> could be because the affine transformation that does the zooming in now
> happening within the Agg backend rather than at the Python level -- i.e.
> there are fewer opportunities for floating-point underflow and the like. Of
> course, SVN trunk is not yet considered stable enough for general usage,
> unfortunately. However, it may be too difficult to fix this in earlier
> versions, because I suspect the fix would involve a fairly significant
> overhaul of where these transformations take place.
Fixing this on the maintenance branch is definitely not a priority.
We've lived with this one for years and it only arises in extreme
zooming. Since it is fixed on the trunk, we can be happy for that.
Thanks,
JDH

Hello
I'd like to know wether Matplotlib can display data with an x axis based on
day, hour, minute and seconds.
I found some example and tutorial that shows that this can be done for year,
month and day but not for hour, minute, seconds.
It looks like matplotlib use the python type date and generate a list of
date within a range (at least in the tutorials).
I'd like to do the same thing but with customized date and time for each of
my points.
So my questions would be : can the plot function display some values with a
certain time (and if possible date) for each one of the values ? Or, can the
plot function use a time object as parameter for x axis ?
I also would like to know wether I can use the same type of functionnality
as in Gnuplot where you can specify the format of data in input and the data
displayed of the x axis. I just guess that the use of the date object
replace any kind of setup in data input but anyway...
Thanks
Regards
Johan Mazel

I was able to reproduce it on 0.90.1 and 0.91.2, but not SVN trunk. 
This could be because the affine transformation that does the zooming in 
now happening within the Agg backend rather than at the Python level -- 
i.e. there are fewer opportunities for floating-point underflow and the 
like. Of course, SVN trunk is not yet considered stable enough for 
general usage, unfortunately. However, it may be too difficult to fix 
this in earlier versions, because I suspect the fix would involve a 
fairly significant overhaul of where these transformations take place.
Cheers,
Mike
John Hunter wrote:
> On Mon, Apr 28, 2008 at 8:21 PM, Sunzen Wang <su...@gm...> wrote:
>
> 
>> > I feel it something to do with zooming in beyond the resolution of the
>> > underlying system (no evidence of this)
>> >
>> I have no idea of it. Could some masters say something about it?
>> 
>
> This is an old, long standing bug. It definitely existed in the agg
> backends in older versions of matplotlib, but I wasn't able to
> reproduce it when I tried on the maintenance branch or the trunk, so I
> don't know if some of Michaels changes have fixed it. It has been a
> while since I looked at it, but the basic problem is that as you zoom
> repeatedly, the logical canvas size (outside the viewport) becomes so
> large that some of the data is overflowing some of aggs
> datastructures. Michaels has done some work to cull points outside
> the viewport at the backend level, so please see if you can reproduce
> this on the svn trunk.
>
> JDH
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
> Don't miss this year's exciting event. There's still time to save 100ドル. 
> Use priority code J8TL2D2. 
> http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
-- 
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA

On Mon, Apr 28, 2008 at 8:21 PM, Sunzen Wang <su...@gm...> wrote:
> > I feel it something to do with zooming in beyond the resolution of the
> > underlying system (no evidence of this)
> >
> I have no idea of it. Could some masters say something about it?
This is an old, long standing bug. It definitely existed in the agg
backends in older versions of matplotlib, but I wasn't able to
reproduce it when I tried on the maintenance branch or the trunk, so I
don't know if some of Michaels changes have fixed it. It has been a
while since I looked at it, but the basic problem is that as you zoom
repeatedly, the logical canvas size (outside the viewport) becomes so
large that some of the data is overflowing some of aggs
datastructures. Michaels has done some work to cull points outside
the viewport at the backend level, so please see if you can reproduce
this on the svn trunk.
JDH

On Tue, Apr 29, 2008 at 2:59 AM, sandric ionut <san...@ya...> wrote:
> I am new to Matplotlib so please excuse me
> I have my data organized in three columns: x,y,z. X and Y represents the
> coordinates of a point and Z is the elevation at that point. I want to
> create a chart that has on y axis the elevation and on x axis the distance
> between the coordinates of the points. The chart displays a line that passes
> all the points, and if is possible interpolate between the elevation points
>
> Can somebody give me an example?
Take a look at this cookbook recipe:
http://www.scipy.org/Cookbook/Matplotlib/Gridding_irregularly_spaced_data
JDH
> Thank you in advance,
>
> Ionut
> ________________________________
> Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try it
> now.
> -------------------------------------------------------------------------
> This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
> Don't miss this year's exciting event. There's still time to save 100ドル.
> Use priority code J8TL2D2.
>
> http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>

On Tue, Apr 29, 2008 at 2:06 AM, Johann Cohen-Tanugi
<co...@sl...> wrote:
> hi John,
> for the record, I have a fedora 8 box, and the shipped version of
> pkgconfig (pkgconfig-0.22-4.fc8) gives me :
>
> [cohen@jarrett matplotlib]$ /usr/bin/pkg-config --libs freetype2
> -lfreetype
>
> No -lz here, if I am not wrong in my querying this tool .....
This looks like the correct incantation, so it appears that your
proplem is a misconfigured freetype2.pc. You might want to post
something to the fedora list, or to the free type list, or simply make
a patch against the pc file. I am not sure who maintains that part.
JDH

On Tue, Apr 29, 2008 at 7:12 AM, Daniel Lidström <dan...@sb...> wrote:
> This is it. The script is a multiplot script, but how to do that part with
> matplotlib
> I think I can work out myself. I am printing two graphs that contain
> "issues" and "tests".
> These values are displayed over time. The data looks like this (both input
> files):
>
> 200711291206 52
> 200711291257 52
> 200711291359 52
I am not a gnuplot user, but I can offer a couple of suggestions.
First, matplotlib has a plotfile command inspired by gnuplot, so you
can use it to plot data from your file. Secondly, recent versions (eg
0.91.2) have transparent support for dates and record arrays, so one
easy solution is to do (assuming your file has headers "date" and
vals" in the example below)
import matplotlib.mlab as mlab
import matplotlib.pyplot as plt
r = mlab.csv2rec('myfile.dat', delimiter=' ')
fig = plt.figure()
ax = fig.add_subplot(111)
ax.plot(r.date, r.vals)
fig.autofmt_xdate()
plt.show()

Hello!
I have recently found out about matplotlib, and at the same time I am having
problems using gnuplot. So I want to convert :-)
However, as I am also quite new to python I have a somewhat steep learning
curve.
It is really only one simple gnuplot script I want to convert, and I have
come
a little bit on the way. Hopefully someone here can provide the last pieces.
Let me show the gnuplot script I am trying to convert:
set terminal png
set title "Historical issue count"
set xlabel "Date"
set xdata time
set timefmt "%Y%m%d%H%M"
set format x "%d/%m"
set ylabel "Issue count"
set xtics nomirror
set ytics nomirror
set border 3
# You should set size before going to multiplot
set size 1.0,1.0
set multiplot
set size 0.5,1
set origin 0,0
plot "< tail -20 last_issues_count_file.txt" using 1:2 title 'Issues' with
linespoints
set title "Number of test cases"
set ylabel "Test cases"
set origin 0.5,0
plot "< tail -20 number_test_cases_file.txt" using 1:2 title 'Tests' with
linespoints
This is it. The script is a multiplot script, but how to do that part with
matplotlib
I think I can work out myself. I am printing two graphs that contain
"issues" and "tests".
These values are displayed over time. The data looks like this (both input
files):
200711291206 52
200711291257 52
200711291359 52
200711291710 52
200711300633 52
200711300916 42
200712030635 42
200712040635 42
Now, I am able to correctly display the second column in matplotlib. Here's
my initial attempt:
# load the issues data
X = load('C:/Users/Daniel.SBG/Desktop/last_issues_count_file.txt')
# select last 50 issues
s = X[:,1][-50:]
# make a nice plot
plot(s, linestyle='-', marker='+', color='red')
This is fine, I can set titles, lables, and legend. But I am missing the
dates on the x-axis.
What I try is this:
# the date format in my input file:
timefmt = '%Y%m%d%H%M'
# create a date formatter (not sure this is the intended purpose)
monthsFmt = DateFormatter(timefmt)
# access the axis
ax = subplot(111)
# set formatter of the axis
ax.xaxis.set_major_formatter(monthsFmt)
This doesn't do anything useful by itself though. I am unsure how to parse
the date column in
my input file and use that as the independent variable of my plots. At the
same time I'd like
to have some nice date formatting on the x-axis.
Any help on these issues are very much appreciated (I don't enjoy gnuplot
anymore...)!
I can probably work from there on.
Thanks in advance!
-- 
Daniel Lidström
SBG AB

On Tue, Apr 29, 2008 at 2:27 AM, G Jones <gle...@gm...> wrote:
> Hello,
> I appologize for a slightly off topic post, but I'm sure someone here
> knows the answer. I like the idea of developing code interactively
> with IPython, but I cannot get past one hurdle. I would like to write
> my code in a module that I then import and instantiate classes or call
> functions from interactively. However, when I try to do this in
> IPython I find that after I import my code, it is cached and even
> after restarting and reimporting, the changes are not recognized. What
> is the paradigm people use to get around this, or what is the
> preferred paradigm for interactive development using IPython.
> Thanks,
> Glenn
While I don't see any persistance across restarts of ipython - I do
within a session.
Check the reload command - which will reimport an already imported
module picking up any fresh changes in the source.
Cheers
Robin

Hello:
 
I am new to Matplotlib so please excuse me
I have my data organized in three columns: x,y,z. X and Y represents the coordinates of a point and Z is the elevation at that point. I want to create a chart that has on y axis the elevation and on x axis the distance between the coordinates of the points. The chart displays a line that passes all the points, and if is possible interpolate between the elevation points
 
Can somebody give me an example?
 
Thank you in advance,
 
Ionut
 ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile. Try it now. http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ 

hi John,
for the record, I have a fedora 8 box, and the shipped version of 
pkgconfig (pkgconfig-0.22-4.fc8) gives me :
[cohen@jarrett matplotlib]$ /usr/bin/pkg-config --libs freetype2
-lfreetype 
No -lz here, if I am not wrong in my querying this tool .....
best,
Johann

2008年4月28日 G Jones <gle...@gm...>:
> Hello,
> I appologize for a slightly off topic post, but I'm sure someone here
> knows the answer.
You're probably right, and people are pretty happy to answer things
here in general. You're more likely to get traction for this question
on the IPython mailing list at <ipython-user at scipy.org>, to which I
have also sent this reply.
> I like the idea of developing code interactively
> with IPython, but I cannot get past one hurdle. I would like to write
> my code in a module that I then import and instantiate classes or call
> functions from interactively.
As far as I can tell, this is a popular and common approach, and one I
use exclusively.
> However, when I try to do this in
> IPython I find that after I import my code, it is cached and even
> after restarting and reimporting, the changes are not recognized. What
> is the paradigm people use to get around this, or what is the
> preferred paradigm for interactive development using IPython.
The method you describe should work without any problem, suggesting to
me that something isn't quite right with your approach yet. Perhaps
you could give a concrete example so we can help debug the problem.
Angus.
-- 
AJC McMorland, PhD candidate
Physiology, University of Auckland
(Nearly) post-doctoral research fellow
Neurobiology, University of Pittsburgh