On Mon, 30 Apr 2012, Vale Fara wrote:
ok, what to do is to generate two sets (x,y) of integer uniform random
numbers so that the following condition is satisfied: the sum of the
numbers obtained in x,y matched two by two (first number obtained in x
with first number obtained in y and so on) is in mean equal to a value
z, mean value that I can decide before the randomization.
Hope this is more clear than before...
It isn't very clear to me. If you generate random X,Y pairs such that
(X+Y)/2=z, then you have a only one random number and a nother that is
completely dependent on it:
X = random
Y = 2z - X.
I'll just tell you one thing you might be able to use, but I don't have
time for this.
To make a vector of N uniformly-distributed random integers in the range
from integer A to integer B, inclusive, you can do this:
floor( runif(N, min=A, max=B+1) )
The floor() function rounds down to the nearest integer. Depending on the
exact nature of the algorithm, it might be possible for B+1 to happen, but
it would be extremely unlikely, if it really is possible.
This should do the same thing:
floor((B-A+1)*runif(N)+A)
The ceiling function can accomplish the same thing. To make random
integers from 1 to K, do this:
ceiling( K*runif(N) )
Mike
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
