[Yt-dev] boundedness change

[Matthew Turk]

> I'm not sure what that energy is.  The 20 minutes I've spent on it
> have thoroughly choked mathematica, and spit out a bunch of log[ a +
> sqrt( b + c ) ] kinds of things.  and the rest of my integration
> skillz are rusty.  So I suspect that there might still be some
> computation for each cell, though not n^2, so might be fast enough to
> do with numpy.  How much faster is a C log than numpy?

If the array is memory aligned, it is a factor of 1.0 faster.

-Matt