matplotlib-users Mailing List for matplotlib

On Tue, 29 Aug 2006, John Hunter wrote:

> I think the PNG corruption happened in the cvs to svn migration. These
> files are from a short course that Fernando and I put together which was a
> general introduction to scientific computing in python, and so it is not
> just matplotlib.  We can probably provide you with a PDF.

John,

   That would be great. When I saw the integrated tutorial
(python->ipython->numpy->scipy->matplotlib) I knew it was the overview I
needed. That's why I tried to build it here.

Thanks,

Rich

-- 
Richard B. Shepard, Ph.D.               |    The Environmental Permitting
Applied Ecosystem Services, Inc.(TM)    |            Accelerator
<http://www.appl-ecosys.com>     Voice: 503-667-4517      Fax: 503-667-8863

>>>>> "Rich" == Rich Shepard <rsh...@ap...> writes:

    Rich>    Having just downloaded the latest code from svn trunk, I
    Rich> started to look at the courses/ directory and was delighted
    Rich> to see all the documentation in .lyx format. However, when I
    Rich> tried to export to pdflatex I discovered that all the .png
    Rich> images in figs/ have damaged headers.

    Rich>    They also do not display or convert.

I think the PNG corruption happened in the cvs to svn migration.
These files are from a short course that Fernando and I put together
which was a general introduction to scientific computing in python,
and so it is not just matplotlib.  We can probably provide you with a
PDF.

JDH

On 8/29/06, PGM <pgm...@gm...> wrote:
> Folks,
> I've just noticed that in order to get the size of a figure in inches, one can
> use:
> >>> fig.get_size_inches()
> However, the corresponding 'set' method is
> >>>fig.set_figsize_inches()
>
> Is this intentional ? Wouldn't it be better to keep the set/get methods
> consistent ?
>
> Thx for your insight

Agreed.  fig.set_figsize seems redundant, so I am changing it to
fig.set_size_inches.  The old one will still be present, but will
issue a deprecation warning.  I'll commit these changes soon.

- Charlie

   Having just downloaded the latest code from svn trunk, I started to look
at the courses/ directory and was delighted to see all the documentation in
.lyx format. However, when I tried to export to pdflatex I discovered that
all the .png images in figs/ have damaged headers.

   They also do not display or convert.

Rich

-- 
Richard B. Shepard, Ph.D.               |    The Environmental Permitting
Applied Ecosystem Services, Inc.(TM)    |            Accelerator
<http://www.appl-ecosys.com>     Voice: 503-667-4517      Fax: 503-667-8863

Folks,
I've just noticed that in order to get the size of a figure in inches, one can 
use:
>>> fig.get_size_inches()
However, the corresponding 'set' method is 
>>>fig.set_figsize_inches()

Is this intentional ? Wouldn't it be better to keep the set/get methods 
consistent ?

Thx for your insight
P.

this could be a very "large" solution depending on the size of your data
sets, but you could make a class for your data and then if you make a
change to the data just import it into your drawing code and then redraw
it.  this way if you have more than 1 set of data you are just creating
more than one instance of your data class.  but if you are using very
large sets of data this could maybe become very slow and ugly.

if you impliment your data class well enough you could make it so that you
have functions in the data class that "add the new stuff" that you need
to.  I'm currently working with a graph that has 6000+ plots of data I am
picking subplots from it and applying a cubic integration to it.  It runs
in a timely manner and i have no complaints.


>Hi,
>Does someone know of a convenient way to save a plotted figure in order to
>reloading it later, possible adding new stuffs to it ? Presently, i keep
>pickling my data and cut&pasting my drawing code, by hand. I'm interested if
>some of you would have better solutions.
>Thanks

>--
>Benjamin Thyreau
>CEA Orsay

Benjamin THYREAU wrote:
> Hi,
> Does someone know of a convenient way to save a plotted figure in order to 
> reloading it later, possible adding new stuffs to it ? Presently, i keep  
> pickling my data and cut&pasting my drawing code, by hand. I'm interested if 
> some of you would have better solutions.

There has been some discussion of making such a mechanism, but it does 
not exist.  I think a better strategy is to encapsulate your code in 
files as functions and/or scripts so as to avoid the cutting and 
pasting.  If initially you are plotting interactively, ipython can make 
it easier for you to do this encapsulation: 
http://ipython.scipy.org/doc/manual/node6.html#SECTION00066000000000000000

Eric

Hi,
Does someone know of a convenient way to save a plotted figure in order to 
reloading it later, possible adding new stuffs to it ? Presently, i keep  
pickling my data and cut&pasting my drawing code, by hand. I'm interested if 
some of you would have better solutions.
Thanks

--
Benjamin Thyreau
CEA Orsay

>>>>> "Samuel" == Samuel GARCIA <sg...@ol...> writes:

    Samuel> Hi all, there is a image.set_clim() but I can't find the
    Samuel> image.get_clim()

    Samuel> Any solution ?

I just committed get_clim to SVN, but in the meantime, you can access
with the image normalization instance, which is what clim is an
interface to anyhow:


In [5]: im = imshow(rand(10,10))

In [6]: im.norm.vmin, im.norm.vmax
Out[6]: (0.011246857233345509, 0.99705302715301514)

Hi all,
there is a image.set_clim()
but I can't find the image.get_clim()

Any solution ?

thank you

Samuel

>>>>> "Tommy" == Tommy Grav <tg...@ma...> writes:

    Tommy> In a plot window (using Tkagg). just right of the menu the
    Tommy> x and y positions of the pointer is shown when the mouse is
    Tommy> in the window. Since I am using small windows this will
    Tommy> change the size of the window. Is there a way to set the
    Tommy> format or remove this text?

This is fixed in mpl svn -- give it a test drive if you can...

JDH

>>>>> "Eric" == Eric Firing <ef...@ha...> writes:

    Eric> I don't think so; and although I have not thought hard about
    Eric> it, my impression is that adding this capability would take
    Eric> a lot of work and redesign.  I expect that in most cases
    Eric> what you describe would not be the desired behavior, so it
    Eric> would have to be a non-default option.


One should be able to write a specific example that exposes this
functionality fairly easily using callback event handling.  I'll
provide a simple example using a line plot to get the approximate
ylimits of the data based on the xaxis zoom limits, and maybe some
enterprising developer can extend this example to an image which sets
the clim based on the xlim and ylim

from pylab import nx, figure, show

t = nx.arange(0.0, 2.0, 0.1)
s = nx.sin(2*nx.pi*t)
fig = figure()
ax = fig.add_subplot(111)
ax.plot(t, s)

def on_xlim(ax):
    xmin, xmax = ax.get_xlim()
    imin, imax = nx.searchsorted(t, (xmin, xmax))
    thiss = s[imin:imax]
    ax.set_ylim(min(thiss), max(thiss))
    ax.figure.canvas.draw()

ax.connect('xlim_changed', on_xlim)
show()

>>>>> "Richard" == Richard Harvey Chapman <hch...@3g...> writes:

    Richard> Briefly, is there a way to get a callback in the event
    Richard> that the canvas has been redrawn (e.g. from a zoom or
    Richard> pan)?

    Richard> matplotlib 0.86.2 python 2.4.1 wxPython 2.6.2.1 Mac OS
    Richard> X.4.7

    Richard> I draw a circle at a certain point. I let the user drag
    Richard> that point to a new location. However, if the plot is
    Richard> redrawn for any reason, I have to redraw my circle in a
    Richard> new position. I can't seem to find any way to get
    Richard> notified of a redraw.

Have you dried connecting to a draw_event?

    def ondraw(event):
        # do something
        pass

   fig.canvas.mpl_connect('draw_event', ondraw)

FYI, the widgets module matplotlib.widgets is a good place to look for
examples of mpl event handling, in addition to the class documentation
in the Event hierarchy in
http://matplotlib.sourceforge.net/matplotlib.backend_bases.html

JDH

>>>>> "Jouni" =3D=3D Jouni K Sepp=E4nen <jk...@ik...> writes:

    Jouni> How I came up with this: I knew that I wanted to make some
    Jouni> of the xticklines invisible, so I looked at the list of
    Jouni> line objects for clues as to what differs between
    Jouni> them. They seem to have xdata and ydata properties, and
    Jouni> ydata is (0,) for half of the lines and (1,) for the other
    Jouni> half, so it looks like it is the vertical position in axis
    Jouni> coordinates. (xdata seems to be in data coordinates.)

Off the top of my head, I didn't remember the answer either.  Here's
how I answered it: I opened up lib/matplotlib/axis.py in my local copy
of the mpl src distro and searched for Tick (you could have done the
same by pointing your browser to
http://matplotlib.sourceforge.net/matplotlib.axis.html#Tick) and saw
the following attributes in the docstring

    Publicly accessible attributes

      tick1line  : a Line2D instance
      tick2line  : a Line2D instance
      gridline   : a Line2D instance
      label1     : a Text instance
      label2     : a Text instance
      gridOn     : a boolean which determines whether to draw the ticklin=
e
      tick1On    : a boolean which determines whether to draw the 1st tic=
kline
      tick2On    : a boolean which determines whether to draw the 2nd tic=
kline
      label1On   : a boolean which determines whether to draw tick label
      label2On   : a boolean which determines whether to draw tick label


I know, and it should be better documented, that for the x-axis, the
tick1 and label1 designations are for the lower axis, and the tick2
and label2 are for the upper (ditto for left/right for the y-axis).
So tick.tick1line controls the properties for the lower tick, etc.

Thus one can do

for tick in ax.xaxis.get_major_ticks():
    tick.tick1line.set_visible(False)
    tick.label2.set_color('darkslategray')

As Jouni notes, the documentation could and should be better, but the
underlying concepts are pretty simple.  A Figure contains multiple
Axes, each of which contains an XAxis and YAxis.  The XAxis and Yaxis
contains XTicks and YTicks, and these contains tick lines (Line2D
instances) and tick labels (Text instances).  Each of these upper-case
thingies is a well-documented matplotlib class (this class containment
hierarchy is also documented in the user's guide).  For future
reference, the links to the class documentation for each of these is

Figure       : http://matplotlib.sourceforge.net/matplotlib.figure.html
Axes/Subplot : http://matplotlib.sourceforge.net/matplotlib.axes.html
XAxis/YAxis  : http://matplotlib.sourceforge.net/matplotlib.axis.html
XTick/YTick  : http://matplotlib.sourceforge.net/matplotlib.axis.html
Line2D       : http://matplotlib.sourceforge.net/matplotlib.lines.html
Text         : http://matplotlib.sourceforge.net/matplotlib.text.html

As Jouni notes, setp and getp are a nice way to get quick interactive
access to the configurable properties, but they are also documented in
the class documentation in the links above.

JDH