CODEWITHSUNDEEP
rmatrixvectorapply
user3640617
user3640617

Reputation: 1576

R: Apply function to matrix with elements of vector as argument

Suppose I want to apply a function to each row of a matrix. One of the function's arguments takes a vector. I would like to apply the first element of the vector to the first row, the second element to the second row, etc.

For example:

set.seed(123)
df<-matrix(runif(100), ncol=10)

var2 <- c(1:10)

MYFUNC <- function(x, Var=NA){ 
  sum(x)/Var 
}

I tried this:

apply(df, 1, function(x) MYFUNC(x, Var=var2))

But that gives me a 10x10 matrix with the function applied to each row & Var combination, whereas I'm only interested in the diagonal elements. I also looked into the mapply function, but I'm not sure how to apply it in this case.

Any help would be really appreciated.

Upvotes: 7

Views: 4148

Answers (2)

Mike H.
Mike H.

Reputation: 14370

Mapply is definitely a possibility. This should work:

mapply(MYFUNC, x = as.data.frame(t(df)), Var = var2)

#V1        V2        V3        V4        V5        V6        V7        V8        V9       V10 
#5.0795111 2.8693537 1.8285747 1.3640238 0.8300597 0.6280441 0.7706310 0.6720132 0.5719003 0.4259674

The issue I think you were running into is that mapply takes either vectors or lists. In R matrices aren't lists, but data.frames are. All you need to do is transpose your matrix and convert to a data.frame and then mapply should work. Each column in a data.frame is an element in the list which is why we have to transpose it (so that each row will be mapped to each element in the vector).

Upvotes: 6

akrun
akrun

Reputation: 887881

As there are two arguments that should be the corresponding rows and elements in matrix/vector respectively, we can loop through the sequence of rows, subset the data and apply the function

sapply(seq_len(nrow(df)), function(i) MYFUNC(df[i,], Var = var2[i]))
#[1] 5.0795111 2.8693537 1.8285747 1.3640238 0.8300597 0.6280441
#[7] 0.7706310 0.6720132 0.5719003 0.4259674

For the specific example, it can be vectorized with rowSums

rowSums(df)/var2
#[1] 5.0795111 2.8693537 1.8285747 1.3640238 0.8300597 0.6280441 
#[7] 0.7706310 0.6720132 0.5719003 0.4259674

Upvotes: 2

Related Questions