Hi, I use R to test a model although I'm not good in using it. To test the model I have to generate random number of arrival rate (lambda) that follows a Gamma distribution. Suppose that in each run I have 5 generated number,>data<-array(1:K*5,c(K,5)); for(i in 1:K) data[i,]=rgamma(5, a,b), that are used in the model. I do not know that how many time, K=?, should I run to achieve at most 95 or 99 percent of interval confidence covering. could you please guide me? best regards, hadi [[alternative HTML version deleted]]