[R] Doubt on benchmarking of R

[Sudip Sinha]

Hi,

In the standard R benchmarking by Simon Urbanek
(http://r.research.att.com/benchmarks/), the trimmed geometric means
(2 extremes eliminated) within the averages in one category has been
calculated. I am unable to figure the rationale behind this, since
this is across different tests (even though they are in the same
category). Help!

Thanks,
Sudip Sinha