matplotlib-users Mailing List for matplotlib

Whenever I run a program that imports matplotlib, it requests pythonw 
rather than python. Here's the message:

This program needs access to the screen. Please run with
'pythonw', not 'python', and only when you are logged in on the main 
display
of your Mac.

Is there any way of getting around this? In particular, when I am 
running python remotely on the terminal, it will not let me use 
matplotlib since access to the screen is required. However, I would 
like to be able to generate plots to file (remotely) without ever 
showing them on the screen. Is there something I need to set in order 
to satisfy this?

Thanks,
--
Christopher J. Fonnesbeck ( c h r i s @ f o n n e s b e c k . o r g )
Georgia Cooperative Fish & Wildlife Research Unit, University of Georgia

>>>>> "Flavio" == Flavio C Coelho <fcc...@fi...> writes:

    Flavio> John, is there a way, in matplotlib, to emulate the
    Flavio> behavior of the matlab "hold" comand to plot multiple
    Flavio> datasets on a single figure?

The default behavior in matplotlib is 'hold on'.

Eg, 

  plot(x1, y1)
  plot(x2, y2)

Both plot to the same figure.

This is one way in which matplotlib differs from matlab.

Thanks for your quick reply.

On 2003.12.16, at 20:47 Japan, John Hunter wrote:

> Is this what you mean?

More precisely, what I mean is like the top-left plot in 
http://www.slac.stanford.edu/grp/ek/hippodraw/canvaswindow.png

If I understand bar() (then hist()) functionalities correctly, it may 
be unable to create such a plot currently without using plot().

from matplotlib.matlab import *
bins = [0, 1, 2, 3]
vals = [3, 4, 5]
x = [bins[0]]
y = [0]
vals.append(0)
for i in range(len(bins) - 1):
    x.append(bins[i])
    y.append(vals[i])
    x.append(bins[i + 1])
    y.append(vals[i])
    x.append(bins[i + 1])
    y.append(vals[i + 1])
plot(x, y)
show()

>>>>> "K" == K KISHIMOTO <ko...@us...> writes:

    K> If possible, I think it is nice for matplotlib to be able to
    K> plot histograms by not only bars but also lines.  My meaning of
    K> "bar histogram plotting" is like the bottom-left plot in
    K> http://jas.freehep.org/images/screenshots/gui2.gif and "line
    K> histogram plotting" is like the others.  There are two reasons.

If I understand you correctly, and from looking at the images you
linked to, all you need to do is set the edge and face properties of
the bars to the same color.  The default edge color is black and the
default face color is blue, so if you want a solid histogram do

    from matplotlib.matlab import *

    mu, sigma = 100, 15
    x = mu + sigma*randn(10000)

    n, bins, patches = hist(x, 200, normed=1)
    set(patches, 'edgecolor', 'b')
    show()

Is this what you mean?

    K> Off course, the bar histogram plotting is more smart in one
    K> case, but in another case the line is better.  In addition, the
    K> support of both bar and line histogram plotting will matplotlib
    K> to be able to have more plotting features that the colors of
    K> line and filled area can be specified separately by the user.

The axes function 'vlines' plots vertical lines.  See the
example/vline_demo.py

Cheers,
John Hunter

Hi,

On 2003.12.15, at 23:09 Japan, John Hunter wrote:

> OK, I'll give it some thought.  Sounds reasonable enough.  It doesn't
> break backwards compatibility, adds useful features, and is more
> consistent with bar.

Thank you so much!

If possible, I think it is nice for matplotlib to be able to plot 
histograms by not only bars but also lines.
My meaning of "bar histogram plotting" is like the bottom-left plot in 
http://jas.freehep.org/images/screenshots/gui2.gif and "line histogram 
plotting" is like the others.
There are two reasons.
(1) Because one bin width becomes a few pixcels, it is hard to see when 
the number of bins is about a few thousand in the bar histogram 
plotting mode.
(2) The color of all of the data filled area become the same by using 
bar and one can specify the line color separately. This brings smart 
view when multiple histograms are plotted like in 
http://root.cern.ch/root/html/examples/gif/hsum.gif

Off course, the bar histogram plotting is more smart in one case, but 
in another case the line is better.
In addition, the support of both bar and line histogram plotting will 
matplotlib to be able to have more plotting features that the colors of 
line and filled area can be specified separately by the user.

It seems that it is relatively not difficult to implement the feature 
by connecting points of [(x0, 0), (x0, y0), (x1, y0), (x1, y1), (x2, 
y1), (x2, y2), ...(and so on)] and will also not break backwards 
compatibility.

If possible, please give a consideration to this.


>>>>>> "K" == K KISHIMOTO <ko...@us...> writes:
>
>     K> I am sorry that my last explanation was insufficient.  The
>     K> point is that I think it is inefficient to execute the
>     K> following two lines in matplotlib.mlab.hist when I know the
>     K> result of "n" already.
>
> Good point ...
>
>     K> Another point is hist() does not support error bar plot (but
>     K> bar() called from hist() does support.)
>
> Another good point....
>
>     K> So, I propose again def hist(x, bins=10, noplot=0, normed=0,
>     K> weights=None, errors=None , **kwargs): in matplotlib.matplab.
>     K> I think if matplotlib supports this by default, it is very
>     K> smart and usefull to many users when making histogram plots
>     K> with matplotlib than letting hist() be mainly for the purpose
>     K> of "calculating" histograms.
>
> OK, I'll give it some thought.  Sounds reasonable enough.  It doesn't
> break backwards compatibility, adds useful features, and is more
> consistent with bar.
>
> JDH
>