A version of Vectorize() for apply() rather than mapply()?

Is there a version of Vectorize that uses apply rather than mapply? I wo开发者_如何转开发uld like to vectorize a function that takes vector input in a way that allows me to pass an array to that function and have a vector returned.

Of course I can write my own wrapper and that works well enough. But I wondered if this functionality is built in?

I can also use Vectorize, but then I have to convert the matrix input to a data.frame for it work. For example:

LAMBDA <- cbind(c(1, .5, .5), c(.5, 1, .5), c(.5, .5, 1))
THETA <- c(0,0,1)
NU <- 21
my.data <- array(1:6, c(3,2))
my.fun <- Vectorize(pmt, vectorize.args="x")

Then

> my.fun(my.data, mean=THETA, S=LAMBDA, df=NU)
[1] 0.4404142 0.8130572 0.9667065 0.9961166 0.9996274 0.9999676

Which is not what I want, but

> my.data <- data.frame(my.data)
> my.fun(my.data, mean=THETA, S=LAMBDA, df=NU)
       X1        X2 
0.8130572 0.9996274 

is what I want.

BTW, pmt is from the mnormt package and is a multivariate student's t cdf.

It seems like you want to define a function my.fun to which you pass in a matrix my.data, and the my.fun will just "know" that it needs to apply the function to each column of my.data. This is what the apply() function is meant to do; you don't need the Vectorize call. So why not just do this:

my.data <- array(1:6, c(3,2))

> apply(my.data, 2, pmt, mean=THETA, S=LAMBDA, df=NU)

[1] 0.8130572 0.9996274