One reason behind the lengthy plot creation times is likely the PDF backend itself.

Whereas the Mac OS X and the Cairo backends make use of new_gc and gc.restore to keep track of the graphics context, the PDF backend uses check_gc and an internal stack of graphics contexts. Since nowadays matplotlib has gc.restore functionality, I don't think that that is needed any more.

See this revision for when gc.restore was added to matplotlib:

http://matplotlib.svn.sourceforge.net/viewvc/matplotlib?view=revision&revision=7112

In the same revision the Mac OS X and Cairo backends were modified to make use of gc.restore. The PDF backend (and the postscript backend also, btw) can be simplified in the same way to speed up these backends, as well as to reduce the output file sizes.

Best,
-Michiel.

--- On Thu, 7/5/12, Gökhan Sever <gokhansever@gmail.com> wrote:

From: Gökhan Sever <gokhansever@gmail.com>
Subject: Re: [Matplotlib-users] Accelerating PDF saved plots
To: "Benjamin Root" <ben.root@ou.edu>
Cc: matplotlib-users@lists.sourceforge.net
Date: Thursday, July 5, 2012, 2:11 PM


38 * 16 = 608
80 / 608 = 0.1316 seconds per plot

At this point, I doubt you are going to get much more speed-ups.  Glad to be of help!

Fabrice -- Good suggestion!  I should have thought of that given how much I use that technique in doing animation.

Ben Root


I am including profiled runs for the records --only first 10 lines to keep e-mail shorter. Total times are longer comparing to the raw run -p executions. I believe profiled run has its own call overhead.

I1 run -p test_speed.py
 171889738 function calls (169109959 primitive calls) in 374.311 seconds

   Ordered by: internal time

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
  4548012   34.583    0.000   34.583    0.000 {numpy.core.multiarray.array}
  1778401   21.012    0.000   46.227    0.000 path.py:86(__init__)
   521816   17.844    0.000   17.844    0.000 artist.py:74(__init__)
  2947090   15.432    0.000   15.432    0.000 weakref.py:243(__init__)
  1778401    9.515    0.000    9.515    0.000 {method 'all' of 'numpy.ndarray' objects}
 13691669    8.654    0.000    8.654    0.000 {getattr}
  1085280    8.550    0.000   17.629    0.000 core.py:2749(_update_from)
  1299904    7.809    0.000   76.060    0.000 markers.py:115(_recache)
       38    7.378    0.194    7.378    0.194 {gc.collect}
 13564851    6.768    0.000    6.768    0.000 {isinstance}




I1 run -p test_speed3.py
 61658708 function calls (60685172 primitive calls) in 100.934 seconds

   Ordered by: internal time

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
   937414    6.638    0.000    6.638    0.000 {numpy.core.multiarray.array}
   374227    4.377    0.000    7.500    0.000 path.py:198(iter_segments)
  6974613    3.866    0.000    3.866    0.000 {getattr}
   542640    3.809    0.000    7.900    0.000 core.py:2749(_update_from)
   141361    3.665    0.000    7.136    0.000 transforms.py:99(invalidate)
324688/161136    2.780    0.000   27.747    0.000 transforms.py:1729(transform)
    64448    2.753    0.000   64.921    0.001 lines.py:463(draw)
   231195    2.748    0.000    7.072    0.000 path.py:86(__init__)
684970/679449    2.679    0.000    3.888    0.000 backend_pdf.py:128(pdfRepr)
    67526    2.651    0.000    7.522    0.000 backend_pdf.py:1226(pathOperations)



--
Gökhan

-----Inline Attachment Follows-----

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/

-----Inline Attachment Follows-----

_______________________________________________
Matplotlib-users mailing list
Matplotlib-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-users