matplotlib-users Mailing List for matplotlib

I've been trying to plot a pcolor over a contourf with a masked array in 
the pcolor so that parts of the contour will show through underneath, 
but whenever I try to do this the pcolor wipes out the contourf.  I can 
do this fine with a contourf over another contourf, but I'm plotting 
model topography, and I would really prefer to leave the discretization 
visible instead of showing contourf's interpolation.  Any way to get a 
pcolor to plot over a contourf without wiping out the contourf beneath it?

Jordan

On 8/31/07, Romain Bignon <ro...@in...> wrote:
> Hello,
>
> I want to get pixels position of a Text object on my imagine, but there isn't
> any methods of this class to get them.
>
> How can I do ?

You can use the t.get_window_extent() method of the text object, with
the caveat that this
only works *after* the canvas has been drawn, so you need to force  a
draw first.  Makre sure you use the same DPI in the figure and savefig
commands if you need these coords for working with hardcopy

from pylab import figure, show

fig = figure(figsize=(6,6), dpi=72)
ax = fig.add_subplot(111)
ax.plot([1,2,3])
t = ax.text(1,2,'hi')
fig.canvas.draw()
left, bottom, width, height = t.get_window_extent().get_bounds()

print left, bottom, width, height

fig.savefig('test', dpi=72)  # make sure you use the same DPI if you save
show()

Hi,

is it somehow possible to have a hatch in parts of the background, which
would achieve something like this pseudo-parameter to axvspan
pylab.axvspan(2, 10,  hatch='//')?

TIA
Christian

OK, I've got it. Previously, I checked the quality of the output image
by two means: by visual inspection in gv and by checking the size of the
output eps images. 
I was puzzled by the different sizes of the images at magnification 1.
Also, convert produces much larger eps files. 

When the size of the output image is set to 6.3246cm (1494px at 600dpi)
and the axes are turned off, both versions appear identical when
printed. 

Thanks for your help,
petr


On Thu, 2007-08-30 at 20:35, Jouni K. Seppänen wrote:
> I don't see a big difference between test-600.eps and test-convert.eps
> when viewed in gv with magnification 10 and 0.1, respectively. Obviously
> there is some resampling in test-600.eps: your source image is 1494 by
> 1494 pixels large, which at 600 dpi is larger than the 5 by 5 cm figure
> created by the script (and the axes are even smaller). test-convert.eps
> has a bounding box of 0 0 1494 1494, so obviously it is a non-resampled
> image at 72 dpi.
> 
> If the problem you are alluding to is in the resampling, perhaps 
> varying the interpolation algorithm will produce a better result? 
> See the docstring of imshow.
> 
> To get a non-resampled image, figimage should work, but it doesn't seem
> to understand PIL images yet...

Hello,

I want to get pixels position of a Text object on my imagine, but there isn't 
any methods of this class to get them.

How can I do ?

Regards,

-- 
Romain Bignon - http://progs.coderz.info

http://www.inl.fr

Hoi,

There is still MPL's polyfit function and I have to admit that Steve
Schmerler's solution looks better that mine, but I've pasted a quick &
dirty solution here:
http://www.python-forum.de/topic-8363.html
It shows the use of polyfit as well as (almost) Steve's approach.

Further examples on linear regression and polynomal regression can be
found in http://matplotlib.sourceforge.net/users_guide_0.90.0.pdf

Also, you might want to have a closer look on the scipy web page:
http://www.scipy.org/ .

Cheers
Christian

Hi - 
I am writing a script that will generate many figures.

Currently this script lives in matlab and outputs these figures to a single .ps file:

if( first )
    print( nfig, '-dpsc2', '-r300', '-loose', filename );
else
    print( nfig, '-dpsc2', '-r300', '-loose', '-append', filename ); 
end

I found the savefig() function.  My question is: does it have something similar to the '-append' flag?  If not, is there another convenient way to save multiple figures to the same .ps file?

Thanks.
martin.

       
---------------------------------
Luggage? GPS? Comic books? 
Check out fitting  gifts for grads at Yahoo! Search.

Petr Danecek <da...@uc...> writes:

> On Fri, 2007-08-24 at 20:03, Jouni K. Seppänen wrote:
>
>> savefig('foo10.ps', dpi=10)
>> savefig('foo100.ps', dpi=100)
>
> In fact, the dpi option does change the resulting PS file, but the
> quality is still very poor - see the example
> 	http://www.ucl.cas.cz/~petr/matplotlib-test.tgz

I don't see a big difference between test-600.eps and test-convert.eps
when viewed in gv with magnification 10 and 0.1, respectively. Obviously
there is some resampling in test-600.eps: your source image is 1494 by
1494 pixels large, which at 600 dpi is larger than the 5 by 5 cm figure
created by the script (and the axes are even smaller). test-convert.eps
has a bounding box of 0 0 1494 1494, so obviously it is a non-resampled
image at 72 dpi.

If the problem you are alluding to is in the resampling, perhaps 
varying the interpolation algorithm will produce a better result? 
See the docstring of imshow.

To get a non-resampled image, figimage should work, but it doesn't seem
to understand PIL images yet...

-- 
Jouni K. Seppänen
http://www.iki.fi/jks

"Alexander Dietz" <Ale...@as...> writes:

> Although the scatter command is called later, any point drawn with the "
> scatter"-command lies behind the black crosses drawn by the "plot" command.
> Any ideas how to get the order right?

Set the zorder property. Try e.g.:

clf()
l=plot(rand(50), rand(50), 'kx')
p=scatter(rand(20), rand(20), 40, c=rand(20), faceted=False)
getp(l[0], 'zorder')
getp(p, 'zorder')
setp(p, zorder=3)

-- 
Jouni K. Seppänen
http://www.iki.fi/jks

Hi,

I have problems with plotting in matplotlib. I want to plot something and
then overlay this plot with a scatter plot, so like this:

clf()
plot( sx, sy, 'kx')
hold(True)
scatter( ix, iy, 40, c=iz, faceted=False)
colorbar()
hold(False)

Although the scatter command is called later, any point drawn with the "
scatter"-command lies behind the black crosses drawn by the "plot" command.
Any ideas how to get the order right?

Cheers
  Alex

Wolfgang Kerzendorf wrote:
> I know this is not completely matplotlib related but perhaps you can 
> help me none the less:
> I want to fit a curve to a set of data. It's a very easy curve: y=ax+b.
> But I want errors for a and b and not only the rms. Is that possible. 
> What tasks do you recommend for doing that.
> Thanks in advance
>     Wolfgang
> 

from http://mathworld.wolfram.com/LeastSquaresFitting.html:

(but here: y = a*x+b, so a <-> b)!

For the standard errors on a and b:

n = float(len(x))
xm = mean(x)
ym = mean(y)
SSxx = dot(x,x) - n*xm**2.0
SSyy = dot(y,y) - n*ym**2.0
SSxy = dot(x,y) - n*xm*ym
r = sqrt(SSxy**2.0 / (SSxx*SSyy))
s = sqrt((SSyy - (SSxy**2.0 / SSxx)) / (n-2.0))

sea = s / sqrt(SSxx)
seb = s * sqrt(1.0/n + (xm**2.0 / SSxx))

The values of sea, seb agree with gnuplot's "Asymptotic Standard Error".

-- 
cheers,
steve

Random number generation is the art of producing pure gibberish as quickly as 
possible.

On 8/30/07, Wolfgang Kerzendorf <wke...@go...> wrote:
> I know this is not completely matplotlib related but perhaps you can
> help me none the less:
> I want to fit a curve to a set of data. It's a very easy curve: y=ax+b.
> But I want errors for a and b and not only the rms. Is that possible.
> What tasks do you recommend for doing that.

gnuplot can do that in a relatively painless way.

I know this is not completely matplotlib related but perhaps you can 
help me none the less:
I want to fit a curve to a set of data. It's a very easy curve: y=ax+b.
But I want errors for a and b and not only the rms. Is that possible. 
What tasks do you recommend for doing that.
Thanks in advance
    Wolfgang

In fact, the dpi option does change the resulting PS file, but the
quality is still very poor - see the example
	http://www.ucl.cas.cz/~petr/matplotlib-test.tgz
pd


On Fri, 2007-08-24 at 20:03, Jouni K. Seppänen wrote:

> I just tried with current svn, and the following script produces two
> results that have visibly different resolutions:
> 
> #!/usr/bin/python
> from pylab import *
> foo = rand(10,10)
> imshow(foo)
> savefig('foo10.ps', dpi=10)
> savefig('foo100.ps', dpi=100)
> 
> Perhaps the original poster could show a bit of code where the scaling
> fails?
> 
> (I'm not sure if figimage is doing the right thing, though...)

nie ninesun wrote:
> 
>   i am new user of matplotlib,i want display two curve on one graph.but i 
> want these two curve have same X axis.but the Y axis is not same...
> 

Seeing as this came up. Is it possible to have more than 2 y axes? For
example:

http://www.unm.edu/~jdfrech/fire-climate.png

The above figure was created by generating 2 figures of the same size in
matplotlib, one with a pair of twinx() y axes and one with a single y axis
then manually splicing them together in Inkscape.
-- 
View this message in context: http://www.nabble.com/multip-axis-tf4344576.html#a12392038
Sent from the matplotlib - users mailing list archive at Nabble.com.

Pierre GM wrote:
> On Wednesday 29 August 2007 10:02:14 John Morgan wrote:
>> I have some data, which I'd like to plot using matplotlib. Some of the data
>> has been flagged by some other software, and I have a boolean array, the
>> same shape as my original data which tells me which data has been flagged

John,

If you are only plotting symbols, you can either use masked arrays, as 
Pierre suggests, or simply plot the selected points:

plot(x[b], y[b], '.', x[~b], y[~b], 'x')

where x, y, and b are all numpy arrays of the same shape.

>> .... 
>> I'm sure I can get this approach to work but I thought I'd ask if there's
>> there a more elegant way to achieve the same aim.
> 
> two words: masked arrays.

In the matplotlib distribution, see examples/masked_demo.py

> 
>>>> import numpy
>>>> flagged = numpy.ma(initial_array, mask=boolean_array)
                numpy.ma.array(...)

>>>> plot(flagged)
> 
> That will take care of the unmasked data.
> For the masked values, just revert the mask.

On Wednesday 29 August 2007 10:02:14 John Morgan wrote:
> I have some data, which I'd like to plot using matplotlib. Some of the data
> has been flagged by some other software, and I have a boolean array, the
> same shape as my original data which tells me which data has been flagged
> .... 
> I'm sure I can get this approach to work but I thought I'd ask if there's
> there a more elegant way to achieve the same aim.

two words: masked arrays.

>>>import numpy
>>>flagged = numpy.ma(initial_array, mask=boolean_array)
>>>plot(flagged)

That will take care of the unmasked data.
For the masked values, just revert the mask.

I have some data, which I'd like to plot using matplotlib. Some of the data
has been flagged by some other software, and I have a boolean array, the
same shape as my original data which tells me which data has been flagged.
What I'd like to do is plot the unflagged data as a single pixels ',' and
flagged data as crosses 'x'.

At the moment I'm plotting all of the data as single pixels, and then
replotting everything as crosses, setting the unflagged data to some
negative number and setting the axes to hide them away off the bottom of the
graph, however this approach is a bit of a pain as you often end up with the
crosses showing up on the bottom of the graph etc.

I'm sure I can get this approach to work but I thought I'd ask if there's
there a more elegant way to achieve the same aim.

Many thanks in advance,

John Morgan

Here's a legend function which allows you to specify loc='outer right'. The
part for 'outer left' doesn't work. It basically wrappes the
pylab.legendmethod, it rescales the plot and sets the location of the
legend to
coordinates outside of the current axes.

def legend(*args, **kwargs):
    """
    Overwrites the pylab legend function.

    It adds another location identfier 'outer right'
    which locates the legend on the right side of the plot

    The args and kwargs are forwarded to the pylab legend function
    """
    if kwargs.has_key('loc'):
        loc = kwargs['loc']
        loc = loc.split()

        if loc[0] == 'outer':
            # make a legend with out the location
            # remove the location setting from the kwargs
            kwargs.pop('loc')
            leg = pylab.legend(loc=(0,0), *args, **kwargs)
            frame = leg.get_frame()
            currentAxes = pylab.gca()
            currentAxesPos = currentAxes.get_position()
            # scale plot by the part which is taken by the legend
            plotScaling = frame.get_width()/currentAxesPos[2]

            if loc[1] == 'right':
                # scale the plot
                currentAxes.set_position((currentAxesPos[0],
currentAxesPos[1],
                                          currentAxesPos[2] *
(1-plotScaling),
                                          currentAxesPos[3]))
                # set x and y coordinates of legend
                leg._loc = (1 + leg.axespad, 1 - frame.get_height())

            # doesn't work
            #if loc[1] == 'left':
            #    # scale the plot
            #    currentAxes.set_position((currentAxesPos[0] +
frame.get_width(),
            #                              currentAxesPos[1],
            #                              currentAxesPos[2] *
(1-plotScaling),
            #                              currentAxesPos[3]))
            #    # set x and y coordinates of legend
            #    leg._loc = (1 -.05 -  leg.axespad - frame.get_width(), 1 -
frame.get_height())

            pylab.draw_if_interactive()
            return leg

    return pylab.legend(*args, **kwargs)


Cheers! Bernhard





On 8/27/07, Wolfgang Kerzendorf <wke...@go...> wrote:
>
> Is there any way to display a legend in a second window or outside the
> plot?
> thanks in advance
>    Wolfgang
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

Hello,

Thanks to your answer.

But now, how can I get Text position on image please ?

Regards,

-- 
Romain Bignon - http://progs.coderz.info

http://www.inl.fr

Hey,

twinx() is what you need.

Download the examples (http://matplotlib.sourceforge.net/=20
matplotlib_examples_0.90.0.zip).  There is an example on how to use =20
twinx() (this is the file two_scales.py)

Keep on Rockin'

Benoit

Le 29-ao=C3=BBt-07 =C3=A0 00:03, nie ninesun a =C3=A9crit :

> hi everyone
>  i am new user of matplotlib,i want display two curve on one =20
> graph.but i want these two curve have same X axis.but the Y axis is =20=

> not same.I read the sample of matplotlib,there is an example about =20
> subplot.but i don't hope to use such one,These two curve should be =20
> displayed in on plot in my application with diffrent axis.Doese =20
> matplotlib have such function? or if there is an demo code about it?
>  thanks
> xz_nie
>
> _________________________________________________________________
> =E4=BA=AB=E7=94=A8=E4=B8=96=E7=95=8C=E4=B8=8A=E6=9C=80=E5=A4=A7=E7=9A=84=
=E7=94=B5=E5=AD=90=E9=82=AE=E4=BB=B6=E7=B3=BB=E7=BB=9F=E2=80=94 MSN =
Hotmail=E3=80=82  http://=20
> www.hotmail.com
>
> ----------------------------------------------------------------------=20=

> ---
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a =20
> browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

--
Dr. Benoit Donnet
Universit=C3=A9 Catholique de Louvain (UCL)
Facult=C3=A9 des Sciences Appliqu=C3=A9es - D=C3=A9partement =
d'Ing=C3=A9nierie =20
Informatique (INGI)
Place Sainte Barbe, 2
1348 Louvain-la-Neuve
Belgium
Phone: +32 10 47 87 18
Home page: http://inl.info.ucl.ac.be/donnet

hi everyone
  i am new user of matplotlib,i want display two curve on one graph.but i 
want these two curve have same X axis.but the Y axis is not same.I read the 
sample of matplotlib,there is an example about subplot.but i don't hope to 
use such one,These two curve should be displayed in on plot in my 
application with diffrent axis.Doese matplotlib have such function? or if 
there is an demo code about it?
  thanks
xz_nie

_________________________________________________________________
享用世界上最大的电子邮件系统― MSN Hotmail。  http://www.hotmail.com  

There were decimal.Decimal object type values in my array. I converted
those to float and all is well now. Don't understand though why the
command line version worked as it was. Hmm... :-)

-Alen

On 8/28/07, Alen Ribic <ale...@gm...> wrote:
> I get the following error when its gets to the line where the bar(...)
> function is called:
>
> "Bbox::update_numerix_xy expected numerix array"
>
> What does this mean? is it referring to the the left, height or width...?
>
> If I execute the same code form the command line it works! But, If I
> call the code from a Web Application then is gives me that error
> above.
>
> I tried via the pylab and matplotlib api. Both give the same error.
>
> Thx
>
> -Alen
>

On 8/28/07, Tom Haddon <to...@gr...> wrote:

> fig.savefig(OUTPUTFILE)

savefig has it's own DPI ( so screen resolution and print resolution
can differ).  So set the figsize in the Figure init method as before,
and then psas dpi to savefig

fig.savefig(blah, dpi=300)

JDH

Hi Folks,

I'm creating a basic graph as follows:

revnos = [ p['revno'] for p in data ]
durations = [ p['duration'] for p in data ]

majorFormatter = FormatStrFormatter('%d')

matplotlib.use('Cairo')
fig = Figure()
canvas = FigureCanvas(fig)
ax = fig.add_subplot(111)
ax.plot(revnos, durations, '-')
ax.set_title('PQM pre-commit hook durations from Revision %s to %s' %
(revnos[0], revnos[-1]))
ax.xaxis.set_major_formatter(majorFormatter)

ax.grid(True)
ax.set_xlabel('Revision')
ax.set_ylabel('Duration in Minutes')
canvas.draw()
fig.savefig(OUTPUTFILE)

Seems to give me a default size of 1200x900 (which I assume is somehow
related to my screen size) - I've tried altering the fig = line as
follows:

fig = Figure(figsize=(8,6), dpi=100)

but can't seem to change the output size of the image.

Thanks, Tom
-- 


----------------------------------
Tom Haddon
mailto:to...@gr...
m +1.415.871.4180
www.greenleaftech.net