[Matplotlib-users] Memory leaks in a web application
Hi,
Following is my post to the Django mailing list from yesterday. The response
was that Django isn't known to leak memory so there should be something off
with matplotlib or rather that way I am using it.
Hopefully someone here could comment on what could be causing the leaks.
Thanks in advance.
I am using matplotlib/pyplot on my site to dynamically generate PNG
plots. And I am experiencing dramatic memory leaks. Within 10-15
hits, my Apache process grows from 15-20M to 100M.
I am using Django 1.0.2-final, Apache 2.2.1, Python 2.4.3, matplotlib
0.98.5.2. The leak happens under both Apache (with mod_wsgi 2.3) and
the development server. My OS is RHEL5.
Below is a simple code snippet that causes the leak. Please let me
know if I am doing something wrong or if there is a better way to
write this. Thanks.
from matplotlib import pyplot
def test_graph (request):
f = pyplot.figure()
ax = f.add_subplot(111)
ax.plot([1,2,3])
ax.fill_between([1,2,3],[1,2,3],[1.1,2.1,3.1])
ax.grid(True)
ax.legend(['hello'],
'upper right', shadow=True, fancybox=True)
ax.set_xlabel('Time')
ax.set_ylabel('Value ')
f.text(.5, 0.93, 'my title', horizontalalignment='center')
response = HttpResponse(content_type='image/png')
### both ways causes a leak
f.savefig( response, format = 'png' )
OR
canvas = FigureCanvas(f)
canvas.print_png(response)
canvas = None
ax = None
f = None
return response
--
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com___
Matplotlib-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/matplotlib-users
Re: [Matplotlib-users] Memory leaks in a web application
A B writes: >f = pyplot.figure() When you do this, matplotlib retains a reference to the figure until you close it so that you can go back to it with e.g. figure(4). So add pyplot.close(f) to your script. Or, even better, use the object-oriented API. To get started with that, see http://matplotlib.sourceforge.net/leftwich_tut.txt -- Jouni K. Seppänen http://www.iki.fi/jks -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Matplotlib-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/matplotlib-users
[Matplotlib-users] controlling tick label size
Trying to find a simple way to shrink the tick labels for this plot -
since I can have many tiny histograms, the labels need to be smaller,
but it isn't obvious to me how to simply shrink them.
code snippet...
num = len(datasets)
fig = plt.figure()
rows = np.int(np.sqrt(num))
cols = np.int(float(num)/float(rows)+.9)
for i in range(num):
ax = fig.add_subplot(rows, cols, i+1)
ax.hist(datasets[i][np.isfinite(datasets[i])], bins=60 )
ax.set_xlabel(labels[i], size = 9)
ax.set_ylabel('Counts', size = 9)
plt.show()
--
---
| Alan K. Jackson| To see a World in a Grain of Sand |
| [email protected] | And a Heaven in a Wild Flower, |
| www.ajackson.org | Hold Infinity in the palm of your hand |
| Houston, Texas | And Eternity in an hour. - Blake |
---
--
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com
___
Matplotlib-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/matplotlib-users
Re: [Matplotlib-users] controlling tick label size
Never mind - I just saw the very timely e-mail from Jouni.
Thanks!
On Sun, 8 Feb 2009 14:16:59 -0600
Alan Jackson wrote:
> Trying to find a simple way to shrink the tick labels for this plot -
> since I can have many tiny histograms, the labels need to be smaller,
> but it isn't obvious to me how to simply shrink them.
>
> code snippet...
>
> num = len(datasets)
> fig = plt.figure()
> rows = np.int(np.sqrt(num))
> cols = np.int(float(num)/float(rows)+.9)
> for i in range(num):
> ax = fig.add_subplot(rows, cols, i+1)
> ax.hist(datasets[i][np.isfinite(datasets[i])], bins=60 )
> ax.set_xlabel(labels[i], size = 9)
> ax.set_ylabel('Counts', size = 9)
> plt.show()
>
>
> --
> ---
> | Alan K. Jackson| To see a World in a Grain of Sand |
> | [email protected] | And a Heaven in a Wild Flower, |
> | www.ajackson.org | Hold Infinity in the palm of your hand |
> | Houston, Texas | And Eternity in an hour. - Blake |
> ---
>
> --
> Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
> software. With Adobe AIR, Ajax developers can use existing skills and code to
> build responsive, highly engaging applications that combine the power of local
> resources and data with the reach of the web. Download the Adobe AIR SDK and
> Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com
> ___
> Matplotlib-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
--
---
| Alan K. Jackson| To see a World in a Grain of Sand |
| [email protected] | And a Heaven in a Wild Flower, |
| www.ajackson.org | Hold Infinity in the palm of your hand |
| Houston, Texas | And Eternity in an hour. - Blake |
---
--
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com
___
Matplotlib-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/matplotlib-users
[Matplotlib-users] release candidate
We have accumulated a number of bug fixes in our stable release branch, so I would like to release the 3rd bugfix release. Please test and report any problems http://matplotlib.sourceforge.net/release-candidates/matplotlib-0.98.5.3.tar.gz Charlie, will you have any time early next week for the windows builds? I can handle the OSX and src release. Thanks, JDH -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Matplotlib-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/matplotlib-users
[Matplotlib-users] different Locator for opposite axes
Hi everyone, I am plotting a figure where I need two independent x axes and two independent y axes. I've tried to use both twinx and twiny at the same time, and this works to some extent, but it looks like it is plotting the labels for the bottom x axis and the right-hand y axis twice, which makes me think that I must be doing something wrong (the numbers appear more 'bold'). The code is below. Is there a better way to do this? In reality, I don't need a different scale for the opposite axes, but I want to specify different Locator functions, but I assume that creating a new axes instance as done below is the only way to do this? Thanks for any advice, Thomas ### fig = figure() ax = fig.add_subplot(111) ax2 = ax.twinx().twiny() for tick in ax.yaxis.get_major_ticks(): tick.label1On = True tick.label2On = False tick.tick1On = True tick.tick2On = False for tick in ax.xaxis.get_major_ticks(): tick.label1On = True tick.label2On = False tick.tick1On = True tick.tick2On = False for tick in ax2.yaxis.get_major_ticks(): tick.label1On = False tick.label2On = True tick.tick1On = False tick.tick2On = True for tick in ax2.xaxis.get_major_ticks(): tick.label1On = False tick.label2On = True tick.tick1On = False tick.tick2On = True ax.scatter([0.4],[0.6]) ax2.scatter([10.],[10.]) draw() ### -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Matplotlib-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/matplotlib-users
Re: [Matplotlib-users] release candidate
Sure. Just send me a note when the src is ready. - Charlie On Sun, Feb 8, 2009 at 3:46 PM, John Hunter wrote: > We have accumulated a number of bug fixes in our stable release > branch, so I would like to release the 3rd bugfix release. Please > test and report any problems > > > http://matplotlib.sourceforge.net/release-candidates/matplotlib-0.98.5.3.tar.gz > > Charlie, will you have any time early next week for the windows > builds? I can handle the OSX and src release. > > Thanks, > JDH > -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Matplotlib-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/matplotlib-users
Re: [Matplotlib-users] rc params strangeness/error
Hi,
Did anyone have any insight on this? Alternatively, anyone know why there
has been little uptake on this question ...(apologies that it was
accidentally sent twice)?
Leo
On Tue, Feb 3, 2009 at 11:42 PM, Leo Trottier wrote:
> Matplotlib 0.98.5.2
> Location: C:\leo\.matplotlib
> Running Windows XP SP2
> Obtained from pythonxy v. 2.1.10
> No rc customizations ...
>
> I've been having a devil of a time getting my rcParams to update
> mid-script. Perhaps this isn't possible, but if that's so, it's not
> entirely clear. Even if it is so, this is seemingly a defect, because in an
> interactive session one expects quite different behavior (and no amount of
> iPython %reset-ing seems to be able to help).
>
> Note that manually setting the font (or what have you) in *title* works
> fine.
>
> Here is some example code:
>
> ## BEGIN #
> import matplotlib
> from matplotlib import rc, rcParams, rcdefaults
> from matplotlib.pyplot import plot, show, figure, title
>
> print matplotlib.__version__
> print matplotlib.get_configdir()
> print
> print "rcParams['font.sans-serif']:", rcParams['font.sans-serif']
> print 'family is:', rcParams['font.family']
> print
> figure(4)
> title('This should be in a sans-serif font')
> show()
>
> rcParams['font.sans-serif'] = rcParams['font.monospace']
> print "rcParams['font.sans-serif']:", rcParams['font.sans-serif']
> print 'family is:', rcParams['font.family']
> print
> figure(1)
> title('This should be in a monospace font')
> show()
>
> rcParams['font.sans-serif'] = rcParams['font.serif']
> print "rcParams['font.sans-serif']:", rcParams['font.sans-serif']
> print 'family is:', rcParams['font.family']
> print
> figure(2)
> title('This should be in a serif font')
> show()
>
> rcdefaults()
> print "rcParams['font.sans-serif']:", rcParams['font.sans-serif']
> print 'family is:', rcParams['font.family']
> print
> figure(3)
> title('This should be back to a sans-serif font')
> show()
> # END
>
--
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com___
Matplotlib-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/matplotlib-users
Re: [Matplotlib-users] different Locator for opposite axes
Since you call twinx then twiny, you're creating two additional axes, not one.
And I guess this is why labels are drawn twice. You may do
def twin(ax):
ax2 = ax.figure.add_axes(ax.get_position(True),
frameon=False)
ax2.yaxis.tick_right()
ax2.yaxis.set_label_position('right')
ax.yaxis.tick_left()
ax2.xaxis.tick_top()
ax2.xaxis.set_label_position('top')
ax.xaxis.tick_bottom()
return ax2
ax = gca()
ax2 = twin(ax)
ax.scatter([0.4],[0.6])
ax2.scatter([10.],[10.])
draw()
Note that you need to manually adjust the view limits of each axes. If
you use sharex or sharey parameters for the axes, you can share their
view limits (this is how axes is created when twinx and twiny is
called). But then you cannot have different tick locators.
In case you need an axes with a same viewlimit as the original one but
just want to place ticks at different position, you may check my
related post.
http://sourceforge.net/mailarchive/forum.php?thread_name=4985DED6.90108%40head.cfa.harvard.edu&forum_name=matplotlib-users
-JJ
On Sun, Feb 8, 2009 at 5:07 PM, Thomas Robitaille
wrote:
> Hi everyone,
>
> I am plotting a figure where I need two independent x axes and two
> independent y axes. I've tried to use both twinx and twiny at the
> same time, and this works to some extent, but it looks like it is
> plotting the labels for the bottom x axis and the right-hand y axis
> twice, which makes me think that I must be doing something wrong (the
> numbers appear more 'bold'). The code is below. Is there a better way
> to do this?
>
> In reality, I don't need a different scale for the opposite axes, but
> I want to specify different Locator functions, but I assume that
> creating a new axes instance as done below is the only way to do this?
>
> Thanks for any advice,
>
> Thomas
>
> ###
> fig = figure()
> ax = fig.add_subplot(111)
> ax2 = ax.twinx().twiny()
> for tick in ax.yaxis.get_major_ticks():
>tick.label1On = True
>tick.label2On = False
>tick.tick1On = True
>tick.tick2On = False
> for tick in ax.xaxis.get_major_ticks():
>tick.label1On = True
>tick.label2On = False
>tick.tick1On = True
>tick.tick2On = False
> for tick in ax2.yaxis.get_major_ticks():
>tick.label1On = False
>tick.label2On = True
>tick.tick1On = False
>tick.tick2On = True
> for tick in ax2.xaxis.get_major_ticks():
>tick.label1On = False
>tick.label2On = True
>tick.tick1On = False
>tick.tick2On = True
> ax.scatter([0.4],[0.6])
> ax2.scatter([10.],[10.])
> draw()
> ###
>
>
> --
> Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
> software. With Adobe AIR, Ajax developers can use existing skills and code to
> build responsive, highly engaging applications that combine the power of local
> resources and data with the reach of the web. Download the Adobe AIR SDK and
> Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com
> ___
> Matplotlib-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
--
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com
___
Matplotlib-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/matplotlib-users
[Matplotlib-users] title pad?
Is there anything like a title pad, similar to the xaxis.LABELPAD? I'd like to see if the plot would look better with the title a bit higher off the plot. Thanks, Che -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Matplotlib-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/matplotlib-users
Re: [Matplotlib-users] title pad?
As far as I know, there is no user settable attribute. But something
like below will work.
ax = gca()
title("My title")
ax.titleOffsetTrans._t = (0., 10.0/72.) # x, y offset in points/72.
default is (0., 5/72.)
ax.titleOffsetTrans.invalidate()
draw()
Alternatively, you may use
ax.title.set_y(1.1) # y position of the title in the normalized axes
coordinate. default 1.0
-JJ
On Mon, Feb 9, 2009 at 2:08 AM, C M wrote:
> Is there anything like a title pad, similar to the xaxis.LABELPAD?
> I'd like to see if the plot would look better with the title a bit higher
> off the plot.
>
> Thanks,
> Che
>
> --
> Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
> software. With Adobe AIR, Ajax developers can use existing skills and code to
> build responsive, highly engaging applications that combine the power of local
> resources and data with the reach of the web. Download the Adobe AIR SDK and
> Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com
> ___
> Matplotlib-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
--
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com
___
Matplotlib-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/matplotlib-users
