On Wed, Sep 28, 2016 at 12:32 AM, Zhilong Liu <lzl200102...@gmail.com> wrote: > Hello all, > > I tried to compare two norm() computations as shown below: > > u = 1.; v = 1.; > > # case 1: constants as arguments > @time for i = 1 : 10000 > g = k * norm([1.; 1.])^3 > end > > # case 2: variables as arguments > @time for i = 1 : 10000 > g = k * norm([u; v])^3 > end > > And the computation speed in case 2 is much slower than case 1. Is there a > way to improve this? Ideally, I would expect them to have the same speed.
Read the performance tip in the manual and don't benchmark in the global scope. > > > > Zhilong Liu > > > > >
