06-29-2012 01:06 AM
I am having problem calculating the lower limit of 99% confidence interval while I am calculating the age standardized rate.
Using below formula (which is one step before getting the CI), somehow it does not compute a value over 126,100.
l_lam99 = gaminv(1 - 0.99) / 2, ncases);
u_lam99 = gaminv(1 + 0.99) / 2, ncases);
For me the ncases are like 129,822 ~ 151,138, which is way higher than 126,100.
Is there any other way to calculate this right or any other solution please?
06-29-2012 07:21 AM
You may want to investigate the QUANTILE function. In SAS 9.3, I had no trouble with the following:
l_lam99 = quantile('GAMMA', (1 - 0.99)/2, ncases);
and inserting 200,000 as the number of cases.