Can I pass row elements as arguments with apply? How? - r

Question

Starting with some sample data:
sample_data <- data.frame(id = 1:3,
x = c(128, 113, 126),
n = c(347, 344, 347),
m = c(335, 334, 347),
index = c(11, 9, -1))
theta <- matrix(c(0.5 ,0.5, 2, 2), nrow=2, ncol=2)
lhs <- function(a, b, g, d, dat){
beta(a + dat$x, b + dat$n - dat$x) / beta(a, b) * beta(g, d + dat$n) / beta(g, d)
}
The function lhs returns a vector of the same number of rows as the argument dat.
rhs <- function(dat, ...){
n = dat$n
m = dat$m
x = dat$x
index = dat$x
temp <- data.frame(i = 0:index,
n = rep(n, index + 1) ,
m = rep(m, index + 1) ,
x = rep(x, index + 1))
sum(beta(a + temp$x, b + temp$m - temp$x + temp$i) / beta(a,b) * beta(g + 1, d + temp$m + temp$i) / beta(g, d))
}
The function rhs works only on a single row because each observation has a different value for index (the index for the sum inside rhs). The intent is to return one value per row in dat. I've tried to do this with apply in the function LL (below).
LL <- function(theta, dat){
a <- theta[1,1]
b <- theta[2,1]
g <- theta[1,2]
d <- theta[2,2]
.lhs <- lhs(a, b, g, d, dat)
.rhs <- ifelse(index > -1, apply(dat, 1, rhs), 0)
sum(log(.lhs+log.rhs))
}
It seems like I need to be able to pass the value of index, n, m, and x to rhs from a given row. That is, not a vector of length(data$n), but the value of n at that row being passed through apply in the function LL.
Is this the correct approach? How can I do such a thing?
Thanks.
Edit
I've clean things up a bit and made a slight modification to the sample data. The correct return value -I think! - can be arrived at by (explicitly passing a,b,g, and d)
sum(lhs(a = theta[1,1],
b = theta[2,1],
g = theta[1,2],
d = theta[2,2],
sample_data)) +
rhs(sample_data[1,]) +
rhs(sample_data[2,]) +
rhs(sample_data[3,])