I have some fully functional code that I'm guessing can be done better/quicker with some savvy R vector tricks; any help to make this run a bit faster would be greatly appreciated; I'm particularly stuck on how to calculate using "row-wise" vectors without iterating explicitly over the dataframe or table ...
library(stats4);
d <- data.frame( ix=c(0,1,2,3,4,5,6,7),
ct=c(253987, 9596, 18680, 2630, 8224, 3590, 5534, 18937),
A=c( 0, 1, 0, 1, 0, 1, 0, 1),
B=c( 0, 0, 1, 1, 0, 0, 1, 1),
C=c( 0, 0, 0, 0, 1, 1, 1, 1)
);
ct <- round(logb(length(d$ix), 2))
ll <- function( th=0.5,
a1=log(0.5), a2=log(0.5), a3=log(0.5),
b1=log(0.5), b2=log(0.5), b3=log(0.5)
) {
a <- exp(sapply(1:ct, function (x) { get(paste("a", x, sep="")) }));
b <- exp(sapply(1:ct, function (x) { get(paste("b", x, sep="")) }));
-sum( d$ct * log( sapply( d$ix,
function (ix, th, a, b) {
x <- d[ix+1,3:(ct+2)]
(th * prod((b ^ (1-x)) * ((1-b) ^ x ))) +
((1-th) * prod((a ^ x ) * ((1-a) ^ (1-x))))
},
th, a, b
)
)
);
}
ml <- mle(ll,
lower=c(0+1e-5, rep(log(0+1e-8), 2*ct)),
upper=c(1-1e-5, rep(log(1-1e-8), 2*ct)),
method="L-BFGS-B"
);For those interested in the math, this is the MLE procedure to estimate the false positive/false negative rates (a and b) of three diagnostic (A, B and C) tests that have the observed performance recapitulated in dataframe "d", but no "gold standard" (sometimes called "latent class analysis", or LCA).
Thanks for any help,
-Aaron
______________________________________________ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
