[GSoC-PortA] Objective Function: mean vs. pamean

[Brian G. Peterson]

On 08/16/2013 07:09 PM, Ross Bennett wrote:
> All,
>
> This isn't critical or important, but thought I would put this out there
> to the group. I was running some of the examples from the 2012 workshop
> script and noticed that any optimization I ran with pamean was really slow.
>
> In commit r2802 I added a testing script
> sandbox/testing_mean_vs_pamean.R that has a simple example comparing the
> two. In my simple benchmark, pamean is about 50 times slower.
> Intuitively I would expect pamean to be slower, but not *that* much slower.
>
> Is this normal? Any idea why this is so slow?


I've never timed it.

It was a simple example of an annualized mean function.

I'm sure it could be easily improved if we took a closer look at it.

Thanks for the heads up.


-- 
Brian G. Peterson
http://braverock.com/brian/
Ph: 773-459-4973
IM: bgpbraverock