On Sat, Nov 3, 2012 at 10:36 PM, Eric Marsden <eric.marsden@free.fr> wrote:
  This is pretty, and hopefully useful to you for tracking performance
  regressions.

Indeed, it is important to realize whether there were regressions and a very important one was just spotted in clos/complex-method
 
  I don't understand the vertical scale on the new plots, though. Is it
  plotting runtime in seconds, like the older (gnuplot?) ones? It does
  seem wierd that ECLi is running faster on many tests than ECL (look at
  CRC40 for instance ...).

My fault: the legend was badly ordered. 

Juanjo

--
Instituto de Física Fundamental, CSIC
c/ Serrano, 113b, Madrid 28006 (Spain) 
http://juanjose.garciaripoll.googlepages.com