matplotlib-users Mailing List for matplotlib (Page 109)

On Aug 15, 2013 1:31 PM, "vwf" <vw...@vu...> wrote:
>
> Got it working!

It would be nice to post your solution here in case others are interested.

>
> (new problems in the making...)
>
>

Good, can't wait to hear them!

Cheers!
Ben Root

After some struggling I got my first plots with quiver working.
A simple plot is very simple, but a complicated one is very different.
Right now I have a 80x80 grid with multiple plots and I plan to go up. 
For this I need full control of the arrow dimensions.

The only way I can get narrow arrow is by setting width:
plt.quiver(x,y,u,v, width=0.001, headwidth=3, scale=0.07,...
works fine for me. But I would like to set the linewidth per arrow.

On stackoverflow I found:
widths = np.linspace(0, 2, X.size)
plt.quiver(X, Y, cos(deg), sin(deg), linewidths=widths)

This did not work for me: the minimum width was too large.
I there a way to create narrow arrows with varying width?

Thanks

Got it working!

(new problems in the making...)

Hello,

I would like to have a set of quiver plots on top of each other: the x/y
positions remain the same but the arrows have different directions. The
arrays are largely empty so the multiple-arrow problem is not that bad
(and very interesting). Can this be done? Right now only the last quiver
plot is visible, the others are gone after plotting.

The objective is multiple arrows on one x/y location, in different
colors. If this can be done in another way please share it.

If it is of any importance: I have arrays with x/y/u/v data, and for
all arrays the x/y positions are the same.

Thanks!

Hi gsal,

Would you mind sharing a simple example how did you manage to that? I am
trying to freeze a simple code with matplotlib for the past two days with no
success, every time I get the "No module backend_tkagg" message... ;(

Many thanks for help in advance!


gsal wrote
> oh, I got it...needed to include backend package in the cf_freeze setup
> script.
> 
> thanks,





--
View this message in context: http://matplotlib.1069221.n5.nabble.com/cx-freeze-ing-tp41780p41808.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

Forgot to mention. I normally use epstopdf to convert eps files to pdf
files and the color information is stripped from the pdf file when embedded
with the color or xcolor packages from within matplotlib. But if you use,
epspdf or epspdftk on linux you can retain color information when
converting eps files to pdf files so can try that as an option.


On Wed, Aug 14, 2013 at 8:50 PM, Xiha <xi...@la...> wrote:

> Thanks Jeffrey. That clarifies why \mathcal works everywhere: it's handled
> by matplotlib's own parser. In contrast, for \textcolor, a call to TeX via
> the lines
>
> from matplotlib import rc
> rc('text', usetex=True)
> rc('text.latex', preamble='\usepackage{color}')
>
> is necessary.
>
> I still don't get \texcolor to work for anything other than ps. This is
> inconvenient as I'm not used to PostScript; for one thing, when I need the
> plot to be 'big' [using plt.figure(figsize=(13.0, 13.0))], the ps file
> created seems to be an A4 format with the plot not fitting onto it: it is
> shown only partially. In contrast, for pdf or svg output, the page size is
> adapted to the figure size -- but in those cases \textcolor does not work...
>
>
> On Wed 14 Aug 2013 03:34:13 CEST, Jeffrey Spencer wrote:
>
>> Have a look here why Mathcal works in all backends:
>>
>> http://matplotlib.org/users/**mathtext.html<http://matplotlib.org/users/mathtext.html>
>>
>> They give an example for an interactive backend which means it would
>> work with any output format in the link you provided. Could also use
>> \textcolor for .pdf output as well since the text rendering would use
>> TeX as well but this wouldn't get you SVG.
>>
>> On Tue, Aug 13, 2013 at 9:00 PM, Xiha <xi...@la...
>> <mailto:xi...@la...>> wrote:
>>
>>     Hello,
>>
>>     I am trying to color-highlight parts of a figure title. I got it
>>     to work via the second ('non-interactive') solution given here
>>     <http://stackoverflow.com/**questions/9169052/partial-**
>> coloring-of-text-in-matplotlib<http://stackoverflow.com/questions/9169052/partial-coloring-of-text-in-matplotlib>
>> **>,
>>
>>     using TeX's \textcolor. It has the advantage (over the first
>>     solution) that you can use .xlabel(), .title() etc. as usual.
>>
>>     However the limitation stated is that it only works when saving
>>     the plot as a PostScript file. I'm finding this to be true: the
>>     coloring does not appear when plotting to the screen rather than
>>     to a file (as with .show()), nor when using matplotlib.use('SVG')
>>     or matplotlib.use('AGG') to get svg or png output (which I would
>>     prefer). This is so even though other 'fancy' TeX commands like
>>     \mathcal do seem to work in all output options.
>>
>>     I am only minimally acquainted with (La)TeX, and fairly new to
>>     Python and matplotlib too, so I don't quite grasp what is going on
>>     here, and whether it is worth digging deeper to try and make it
>>     work. So: why is there a difference in success between using
>>     (e.g.) \mathcal versus \textcolor over different output options?
>>
>>     Many thanks!
>>     ||
>>
>>     ------------------------------**------------------------------**
>> ------------------
>>     Get 100% visibility into Java/.NET code with AppDynamics Lite!
>>     It's a free troubleshooting tool designed for production.
>>     Get down to code-level detail for bottlenecks, with <2% overhead.
>>     Download for free and get started troubleshooting in minutes.
>>     http://pubads.g.doubleclick.**net/gampad/clk?id=48897031&iu=**
>> /4140/ostg.clktrk<http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk>
>>     ______________________________**_________________
>>     Matplotlib-users mailing list
>>     Matplotlib-users@lists.**sourceforge.net<Mat...@li...>
>>     <mailto:Matplotlib-users@**lists.sourceforge.net<Mat...@li...>
>> >
>>     https://lists.sourceforge.net/**lists/listinfo/matplotlib-**users<https://lists.sourceforge.net/lists/listinfo/matplotlib-users>
>>
>>
>>
>
>

The .tex parser for matplotlib uses dvi files that when converted to be
used in a pdf or other backends lose their color information is my guess.
If you output a ps file (or an eps file which you might want to do so you
don't have to set the page size). Then run epstopdf or pstopdf the color is
striped from the file as well. I don't know enough about the actual
mechanisms to provide much more information then this but hope that helps.


On Wed, Aug 14, 2013 at 8:50 PM, Xiha <xi...@la...> wrote:

> Thanks Jeffrey. That clarifies why \mathcal works everywhere: it's handled
> by matplotlib's own parser. In contrast, for \textcolor, a call to TeX via
> the lines
>
> from matplotlib import rc
> rc('text', usetex=True)
> rc('text.latex', preamble='\usepackage{color}')
>
> is necessary.
>
> I still don't get \texcolor to work for anything other than ps. This is
> inconvenient as I'm not used to PostScript; for one thing, when I need the
> plot to be 'big' [using plt.figure(figsize=(13.0, 13.0))], the ps file
> created seems to be an A4 format with the plot not fitting onto it: it is
> shown only partially. In contrast, for pdf or svg output, the page size is
> adapted to the figure size -- but in those cases \textcolor does not work...
>
>
> On Wed 14 Aug 2013 03:34:13 CEST, Jeffrey Spencer wrote:
>
>> Have a look here why Mathcal works in all backends:
>>
>> http://matplotlib.org/users/**mathtext.html<http://matplotlib.org/users/mathtext.html>
>>
>> They give an example for an interactive backend which means it would
>> work with any output format in the link you provided. Could also use
>> \textcolor for .pdf output as well since the text rendering would use
>> TeX as well but this wouldn't get you SVG.
>>
>> On Tue, Aug 13, 2013 at 9:00 PM, Xiha <xi...@la...
>> <mailto:xi...@la...>> wrote:
>>
>>     Hello,
>>
>>     I am trying to color-highlight parts of a figure title. I got it
>>     to work via the second ('non-interactive') solution given here
>>     <http://stackoverflow.com/**questions/9169052/partial-**
>> coloring-of-text-in-matplotlib<http://stackoverflow.com/questions/9169052/partial-coloring-of-text-in-matplotlib>
>> **>,
>>
>>     using TeX's \textcolor. It has the advantage (over the first
>>     solution) that you can use .xlabel(), .title() etc. as usual.
>>
>>     However the limitation stated is that it only works when saving
>>     the plot as a PostScript file. I'm finding this to be true: the
>>     coloring does not appear when plotting to the screen rather than
>>     to a file (as with .show()), nor when using matplotlib.use('SVG')
>>     or matplotlib.use('AGG') to get svg or png output (which I would
>>     prefer). This is so even though other 'fancy' TeX commands like
>>     \mathcal do seem to work in all output options.
>>
>>     I am only minimally acquainted with (La)TeX, and fairly new to
>>     Python and matplotlib too, so I don't quite grasp what is going on
>>     here, and whether it is worth digging deeper to try and make it
>>     work. So: why is there a difference in success between using
>>     (e.g.) \mathcal versus \textcolor over different output options?
>>
>>     Many thanks!
>>     ||
>>
>>     ------------------------------**------------------------------**
>> ------------------
>>     Get 100% visibility into Java/.NET code with AppDynamics Lite!
>>     It's a free troubleshooting tool designed for production.
>>     Get down to code-level detail for bottlenecks, with <2% overhead.
>>     Download for free and get started troubleshooting in minutes.
>>     http://pubads.g.doubleclick.**net/gampad/clk?id=48897031&iu=**
>> /4140/ostg.clktrk<http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk>
>>     ______________________________**_________________
>>     Matplotlib-users mailing list
>>     Matplotlib-users@lists.**sourceforge.net<Mat...@li...>
>>     <mailto:Matplotlib-users@**lists.sourceforge.net<Mat...@li...>
>> >
>>     https://lists.sourceforge.net/**lists/listinfo/matplotlib-**users<https://lists.sourceforge.net/lists/listinfo/matplotlib-users>
>>
>>
>>
>
>

Thanks Jeffrey. That clarifies why \mathcal works everywhere: it's 
handled by matplotlib's own parser. In contrast, for \textcolor, a call 
to TeX via the lines

from matplotlib import rc
rc('text', usetex=True)
rc('text.latex', preamble='\usepackage{color}')

is necessary.

I still don't get \texcolor to work for anything other than ps. This is 
inconvenient as I'm not used to PostScript; for one thing, when I need 
the plot to be 'big' [using plt.figure(figsize=(13.0, 13.0))], the ps 
file created seems to be an A4 format with the plot not fitting onto 
it: it is shown only partially. In contrast, for pdf or svg output, the 
page size is adapted to the figure size -- but in those cases 
\textcolor does not work...

On Wed 14 Aug 2013 03:34:13 CEST, Jeffrey Spencer wrote:
> Have a look here why Mathcal works in all backends:
>
> http://matplotlib.org/users/mathtext.html
>
> They give an example for an interactive backend which means it would
> work with any output format in the link you provided. Could also use
> \textcolor for .pdf output as well since the text rendering would use
> TeX as well but this wouldn't get you SVG.
>
> On Tue, Aug 13, 2013 at 9:00 PM, Xiha <xi...@la...
> <mailto:xi...@la...>> wrote:
>
>     Hello,
>
>     I am trying to color-highlight parts of a figure title. I got it
>     to work via the second ('non-interactive') solution given here
>     <http://stackoverflow.com/questions/9169052/partial-coloring-of-text-in-matplotlib>,
>     using TeX's \textcolor. It has the advantage (over the first
>     solution) that you can use .xlabel(), .title() etc. as usual.
>
>     However the limitation stated is that it only works when saving
>     the plot as a PostScript file. I'm finding this to be true: the
>     coloring does not appear when plotting to the screen rather than
>     to a file (as with .show()), nor when using matplotlib.use('SVG')
>     or matplotlib.use('AGG') to get svg or png output (which I would
>     prefer). This is so even though other 'fancy' TeX commands like
>     \mathcal do seem to work in all output options.
>
>     I am only minimally acquainted with (La)TeX, and fairly new to
>     Python and matplotlib too, so I don't quite grasp what is going on
>     here, and whether it is worth digging deeper to try and make it
>     work. So: why is there a difference in success between using
>     (e.g.) \mathcal versus \textcolor over different output options?
>
>     Many thanks!
>     ||
>
>     ------------------------------------------------------------------------------
>     Get 100% visibility into Java/.NET code with AppDynamics Lite!
>     It's a free troubleshooting tool designed for production.
>     Get down to code-level detail for bottlenecks, with <2% overhead.
>     Download for free and get started troubleshooting in minutes.
>     http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
>     _______________________________________________
>     Matplotlib-users mailing list
>     Mat...@li...
>     <mailto:Mat...@li...>
>     https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>

Have a look here why Mathcal works in all backends:

http://matplotlib.org/users/mathtext.html

They give an example for an interactive backend which means it would work
with any output format in the link you provided. Could also use \textcolor
for .pdf output as well since the text rendering would use TeX as well but
this wouldn't get you SVG.

On Tue, Aug 13, 2013 at 9:00 PM, Xiha <xi...@la...> wrote:

>  Hello,
>
> I am trying to color-highlight parts of a figure title. I got it to work
> via the second ('non-interactive') solution given here<http://stackoverflow.com/questions/9169052/partial-coloring-of-text-in-matplotlib>,
> using TeX's \textcolor. It has the advantage (over the first solution) that
> you can use .xlabel(), .title() etc. as usual.
>
> However the limitation stated is that it only works when saving the plot
> as a PostScript file. I'm finding this to be true: the coloring does not
> appear when plotting to the screen rather than to a file (as with .show()),
> nor when using matplotlib.use('SVG') or matplotlib.use('AGG') to get svg or
> png output (which I would prefer). This is so even though other 'fancy' TeX
> commands like \mathcal do seem to work in all output options.
>
> I am only minimally acquainted with (La)TeX, and fairly new to Python and
> matplotlib too, so I don't quite grasp what is going on here, and whether
> it is worth digging deeper to try and make it work. So: why is there a
> difference in success between using (e.g.) \mathcal versus \textcolor over
> different output options?
>
> Many thanks!
>
>
> ------------------------------------------------------------------------------
> Get 100% visibility into Java/.NET code with AppDynamics Lite!
> It's a free troubleshooting tool designed for production.
> Get down to code-level detail for bottlenecks, with <2% overhead.
> Download for free and get started troubleshooting in minutes.
> http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>

Hello,

I am trying to color-highlight parts of a figure title. I got it to work 
via the second ('non-interactive') solution given here 
<http://stackoverflow.com/questions/9169052/partial-coloring-of-text-in-matplotlib>, 
using TeX's \textcolor. It has the advantage (over the first solution) 
that you can use .xlabel(), .title() etc. as usual.

However the limitation stated is that it only works when saving the plot 
as a PostScript file. I'm finding this to be true: the coloring does not 
appear when plotting to the screen rather than to a file (as with 
.show()), nor when using matplotlib.use('SVG') or matplotlib.use('AGG') 
to get svg or png output (which I would prefer). This is so even though 
other 'fancy' TeX commands like \mathcal do seem to work in all output 
options.

I am only minimally acquainted with (La)TeX, and fairly new to Python 
and matplotlib too, so I don't quite grasp what is going on here, and 
whether it is worth digging deeper to try and make it work. So: why is 
there a difference in success between using (e.g.) \mathcal versus 
\textcolor over different output options?

Many thanks!
||

As I'm researching what we may want to do for better continuous 
integration, I'm remembering that at least one person, Thomas Kluyver, 
is producing daily automated builds (for Ubuntu) here:

https://launchpad.net/~takluyver/+archive/matplotlib-daily 
<https://launchpad.net/%7Etakluyver/+archive/matplotlib-daily>

Is anyone else out there doing anything similar for other Linux distros 
or other platforms?  a) I'd like to list these things on the main 
website, and b) I'd like to look at how these kinds of things might make 
sense as part of a broader CI strategy.

Cheers,
Mike

On Tue, Aug 13, 2013 at 04:43:19PM +1000, Stephen Gibson wrote:
> Call   'figure()'   for each plot.
Like this you mean?

import matplotlib.pyplot as plt
a=plt.figure()
a=plt.plot([1, 2], [1, 2])
plt.savefig('1.png', dpi=100)
a=plt.figure()
a=plt.plot([1, 2], [2,1])
plt.savefig('2.png', dpi=100)

Thank you!

On 2013/08/12 8:35 PM, vwf wrote:
> Thank you for you reply. I tried to create one after the other but when
> I did this my second plot was on top of the first one. The old plot
> needs to be "flushed" before starting the second one.
>
> This doesn't work:
> import matplotlib.pyplot as plt
> a=plt.plot([1, 2], [1, 2])
> plt.savefig('1.png', dpi=100)

plt.close()


> b=plt.plot([1, 2], [2,1])
> plt.savefig('2.png', dpi=100)
>
> In 2.png, a and b are on top of each other
>

Call   'figure()'   for each plot.

see:  http://matplotlib.org/api/pyplot_api.html

matplotlib.pyplot.figure(/num=None/, /figsize=None/, /dpi=None/, 
/facecolor=None/, /edgecolor=None/, /frameon=True/, /FigureClass=<class 
'matplotlib.figure.Figure'>/, /**kwargs/)

    Creates a new figure.

    Parameters : 	

    *num* : integer or string, optional, default: none

        If not provided, a new figure will be created, and a the figure
        number will be increamted. The figure objects holds this number
        in a number attribute. If num is provided, and a figure with
        this id already exists, make it active, and returns a reference
        to it. If this figure does not exists, create it and returns it.
        If num is a string, the window title will be set to this
        figure's num.



Steve.

Thank you for you reply. I tried to create one after the other but when
I did this my second plot was on top of the first one. The old plot
needs to be "flushed" before starting the second one.

This doesn't work:
import matplotlib.pyplot as plt
a=plt.plot([1, 2], [1, 2])
plt.savefig('1.png', dpi=100)
b=plt.plot([1, 2], [2,1])
plt.savefig('2.png', dpi=100)

In 2.png, a and b are on top of each other


On Tue, Aug 13, 2013 at 12:10:32AM -0600, Joseph Hardin wrote:
> Can you provide us with more information? You can create one plot, save it,
> and then create the second, or is there something more specific you are
> looking for?
> 
> 
> On Mon, Aug 12, 2013 at 11:58 PM, vwf <vw...@vu...> wrote:
> 
> > Hello,
> >
> > I need to create two plots (png files) in one go, two unrelated views of
> > the same dataset. There is good documentation about subplots but I
> > cannot locate documentation about two plots. Can someone tell me how it
> > is done?
> >
> >
> > ------------------------------------------------------------------------------
> > Get 100% visibility into Java/.NET code with AppDynamics Lite!
> > It's a free troubleshooting tool designed for production.
> > Get down to code-level detail for bottlenecks, with <2% overhead.
> > Download for free and get started troubleshooting in minutes.
> > http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
> > _______________________________________________
> > Matplotlib-users mailing list
> > Mat...@li...
> > https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> >
> 
> 
> 
> -- 
> Joseph Hardin, MSEE
> Colorado State University
> Radar and Communications Laboratory
> One must do not violence to nature, nor model it in conformity to any
> blindly formed chimaera.
> -Janos Boylai

Can you provide us with more information? You can create one plot, save it,
and then create the second, or is there something more specific you are
looking for?


On Mon, Aug 12, 2013 at 11:58 PM, vwf <vw...@vu...> wrote:

> Hello,
>
> I need to create two plots (png files) in one go, two unrelated views of
> the same dataset. There is good documentation about subplots but I
> cannot locate documentation about two plots. Can someone tell me how it
> is done?
>
>
> ------------------------------------------------------------------------------
> Get 100% visibility into Java/.NET code with AppDynamics Lite!
> It's a free troubleshooting tool designed for production.
> Get down to code-level detail for bottlenecks, with <2% overhead.
> Download for free and get started troubleshooting in minutes.
> http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>



-- 
Joseph Hardin, MSEE
Colorado State University
Radar and Communications Laboratory
One must do not violence to nature, nor model it in conformity to any
blindly formed chimaera.
-Janos Boylai

Hello,

I need to create two plots (png files) in one go, two unrelated views of
the same dataset. There is good documentation about subplots but I
cannot locate documentation about two plots. Can someone tell me how it
is done?

On Mon, Aug 12, 2013 at 7:01 AM, Michael Droettboom <md...@st...> wrote:
>  I propose to fix this by turning on interactive only when
> running at an interactive console.

I embed MPL more than other uses, and this sounds like a fine solution to me/

Thanks,

-Chris


-- 

Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/OR&R            (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chr...@no...

I'm considering changing the behavior of the rcParam |interactive| 
(which also can be set through |matplotlib.interactive()| and 
|pyplot.ion()| and |pyplot.ioff()|). Currently, when setting 
|interactive| to |True|, running any sort of matplotlib plot as a script 
will fail to display a window. This can be very surprising if a user 
turns on |interactive| because they prefer its behavior in IPython, but 
are then surprised that none of their scripts continue to work. I 
propose to fix this by turning on |interactive| only when running at an 
interactive console.

See the pull request http://github.com/matplotlib/matplotlib/pull/2286 
for an implementation.

I'm trying to rule out any negative impact of this change, and I would 
appreciate any feedback if this change will have a negative impact on 
your application.

Mike

oh, I got it...needed to include backend package in the cf_freeze setup
script.

thanks,



--
View this message in context: http://matplotlib.1069221.n5.nabble.com/cx-freeze-ing-tp41780p41782.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

Hhhhmmm...what exactly do you have in mind? Sure, many of the steps to plot
something are the same; to be sure, the data changes and possibly the type
of plot...

what kind of application are you writing?
how much freedom do you really want to give to the user?
is the plot specific enough? given the context of the application?
is the user expected to know python and matplotlib?

have you tried eval?

have you thought about simply providing a python console along with your
application?

...just throwing questions at you to make you think more about this.



--
View this message in context: http://matplotlib.1069221.n5.nabble.com/Newbie-question-executing-commands-from-a-string-tp41756p41781.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

So, I have a working program that starts like this:

import os
import sys
import numpy as np
from pylab import *
from matplotlib.widgets import Slider
from matplotlib.patches import FancyBboxPatch, Rectangle

but when I cx_freeze it, it complains about some missing backend_ stuff, so,
I do:

import os
import sys
import numpy as np
*import matplotlib
matplotlib.use("TkAgg")
*from pylab import *
from matplotlib.widgets import Slider
from matplotlib.patches import FancyBboxPatch, Rectangle

but the problem continues...

...anybody know how to solve this problem?

I would appreciate very much.

Oh, here is the setup file generated with cx_freeze


from cx_Freeze import setup, Executable

# Dependencies are automatically detected, but it might need
# fine tuning.
buildOptions = dict(packages = [], excludes = [])

import sys
base = 'Win32GUI' if sys.platform=='win32' else None

executables = [
    Executable('c.py', base=base)
]

setup(name='Cross-slot-flux',
      version = '1.0',
      description = 'Visual',
      options = dict(build_exe = buildOptions),
      executables = executables)






--
View this message in context: http://matplotlib.1069221.n5.nabble.com/cx-freeze-ing-tp41780.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

On Fri, Aug 9, 2013 at 11:15 AM, Martin Mokrejs <mmo...@fo...
> wrote:

> Hi Ben,
>   thank your for your comments. OK, here are revised patches. I see a hot
> spot
> in artist.py where the getattr() calls are too expensive. Actually, those
> under
> the callable() path.
>
>
Ah, yes... one of the biggest warts in our codebase. Does get_aliases()
really get called that often?  If so, I doubt changes with regards to
startwith() is the best use of our time.  I would imagine that a refactor
that caches possible aliases. Hell, I would just as soon like to see the
entire aliases framework ripped out and redone more correctly in the first
place.

As for the other startswith() changes, there are some subtle differences in
the change that has to be considered. First, is there a guarantee that the
string being indexed is not empty? startswith() would handle that
correctly, while indexing would throw an exception (and setting up code to
try...catch those exceptions would reduce readability and probably reduce
performance back to where we started).

I think that the *biggest* improvement we are going to get is from your
patch to figure.py because it touches on some very deep code that is
executed very frequently, and we were doing in probably the most
inefficient manner.

Again, I really stress the importance of setting up a github account and
submitting a PR. Once you do that, you are all set to go for contributing
to many other great projects (numpy, scipy, ipython, etc.).

Cheers!
Ben Root

Hi Ben,
  thank your for your comments. OK, here are revised patches. I see a hot spot
in artist.py where the getattr() calls are too expensive. Actually, those under
the callable() path.


 793 class ArtistInspector:

 817     def get_aliases(self):
 818         """
 819         Get a dict mapping *fullname* -> *alias* for each *alias* in
 820         the :class:`~matplotlib.artist.ArtistInspector`.
 821 
 822         Eg., for lines::
 823 
 824           {'markerfacecolor': 'mfc',
 825            'linewidth'      : 'lw',
 826           }
 827 
 828         """
 829         names = [name for name in dir(self.o) if
 830                  (name[:4] in ['set_', 'get_'])
 831                  and callable(getattr(self.o, name))]
 832         aliases = {}
 833         for name in names:
 834             func = getattr(self.o, name)
 835             if not self.is_alias(func):
 836                 continue
 837             docstring = func.__doc__
 838             fullname = docstring[10:]
 839             aliases.setdefault(fullname[4:], {})[name[4:]] = None
 840         return aliases


Another hot spot is setp() artist.py, actually its get_aliases() on line 817,
which again leads to getattr() and callable(). The problem is there are millions of
their calls.


So, once again, my all my patches attached (figure.py.patch artist.py.patch have changed).
Martin



Benjamin Root wrote:
> 
> On Fri, Aug 9, 2013 at 9:04 AM, Martin Mokrejs <mmo...@fo... <mailto:mmo...@fo...>> wrote:
> 
>     Hi Phil,
> 
>     Phil Elson wrote:
>     > Hi Martin,
>     >
>     > Thanks for this - we are really interested in speeding up the scatter and barchart plotting with large data sets. In fact, we've done some work (https://github.com/matplotlib/matplotlib/pull/2156) recently to make the situation better.
>     >
>     > I'd really like to review these changes (against matplotlib master), and the best possible solution to doing this is if you were to submit a pull request. If the changes you have made are logically seperable, then I'd encourage you to make a few PRs, but otherwise, a single PR with all of these changes would be great.
> 
>     I went through the changes there and they just cope with other pieces of matplotlib.
>     My changes are general python improvements moving away from str.startswith()
>     and using generators instead of for loops. Just apply the patches yourself and see.
>     ;)
> 
>     >
>     > Would you mind turning these patches into PR(s)? (https://github.com/matplotlib/matplotlib/compare/)
> 
>     Um, I don't know what to do on that page, sorry. I don't see how to upload my patch file or patched file
>     to be compared with master. :(
> 
>     >
>     > Thanks!
> 
>     I am sorry but I just don't have time to fiddle with github. It is just awkward. I even failed to download
>     diffs of the changes from https://github.com/matplotlib/matplotlib/pull/2156/commits.
> 
>     I rather continue studying runsnake output. ;-)
> 
>     Martin
> 
> 
> A snippet from one of you patches:
> 
>          dsu = []
>  
> -        for a in self.patches:
> -            dsu.append( (a.get_zorder(), a, a.draw, [renderer]))
> +        [ dsu.append( (x.get_zorder(), x, x.draw, [renderer])) for x in self.patches ]
> 
> Yes, we certainly should use list-comprehensions here, but you are using it incorrectly.  It should be:
> 
> dsu = [(x.get_zorder(), x, x.draw, [renderer])) for x in self.patches ]
> 
> And then, further down, do the following:
> 
> dsu.extend((x.get_zorder(), x, x.draw, [renderer])) for x in self.lines)
> 
> Note the generator form of the comprehension as opposed to the list comprehension form. List comprehensions should *always* be assigned to something. List comprehensions should only be for replacing the for-append idiom in python.
> 
> Thank you though for pointing out parts of the code that can benefit from revisions. I certainly hope you can get this put together as a pull request on github so we can work to make this patch better!
> 
> Ben Root
> 
> P.S. - I <3 runsnakerun!

On Fri, Aug 9, 2013 at 9:04 AM, Martin Mokrejs
<mmo...@fo...>wrote:

> Hi Phil,
>
> Phil Elson wrote:
> > Hi Martin,
> >
> > Thanks for this - we are really interested in speeding up the scatter
> and barchart plotting with large data sets. In fact, we've done some work (
> https://github.com/matplotlib/matplotlib/pull/2156) recently to make the
> situation better.
> >
> > I'd really like to review these changes (against matplotlib master), and
> the best possible solution to doing this is if you were to submit a pull
> request. If the changes you have made are logically seperable, then I'd
> encourage you to make a few PRs, but otherwise, a single PR with all of
> these changes would be great.
>
> I went through the changes there and they just cope with other pieces of
> matplotlib.
> My changes are general python improvements moving away from
> str.startswith()
> and using generators instead of for loops. Just apply the patches yourself
> and see.
> ;)
>
> >
> > Would you mind turning these patches into PR(s)? (
> https://github.com/matplotlib/matplotlib/compare/)
>
> Um, I don't know what to do on that page, sorry. I don't see how to upload
> my patch file or patched file
> to be compared with master. :(
>
> >
> > Thanks!
>
> I am sorry but I just don't have time to fiddle with github. It is just
> awkward. I even failed to download
> diffs of the changes from
> https://github.com/matplotlib/matplotlib/pull/2156/commits.
>
> I rather continue studying runsnake output. ;-)
>
> Martin
>
>
A snippet from one of you patches:

         dsu = []

-        for a in self.patches:
-            dsu.append( (a.get_zorder(), a, a.draw, [renderer]))
+        [ dsu.append( (x.get_zorder(), x, x.draw, [renderer])) for x in
self.patches ]

Yes, we certainly should use list-comprehensions here, but you are using it
incorrectly.  It should be:

dsu = [(x.get_zorder(), x, x.draw, [renderer])) for x in self.patches ]

And then, further down, do the following:

dsu.extend((x.get_zorder(), x, x.draw, [renderer])) for x in self.lines)

Note the generator form of the comprehension as opposed to the list
comprehension form. List comprehensions should *always* be assigned to
something. List comprehensions should only be for replacing the for-append
idiom in python.

Thank you though for pointing out parts of the code that can benefit from
revisions. I certainly hope you can get this put together as a pull request
on github so we can work to make this patch better!

Ben Root

P.S. - I <3 runsnakerun!