Most of the computation time is in the functions qvnorm. You can win a
little bit by optimizing the code, but the gain is relatively small.
You can also decrease the interval used to evaluate qvnorm to win some
speed there. As you look for the upper tail, no need to evaluate the
function in negativ
use Rprof to determine where your function is spending time.
What is the problem you are trying to solve?
Sent from my iPhone.
On Jun 23, 2010, at 5:21, li li wrote:
Dear all,
I have the following program for a multiple comparison procedure.
There are two functions for the two steps. First
2 matches
Mail list logo
