Avoiding a loop in R

Question

I am trying to avoid a loop in R. But I am having difficulties. At the moment this is the code that I have but I would really like to know if there is a direct way to do it (no loop). Triangle is just a IxJ matrix

I = 10
J = 10
xi <- rep(0,J) 
xi[1] <- 1
for(j  in 2:J){
  xi[j]<-(1/(I-j+1))*sum(log(Triangle[1:(I-j+1),j]/Triangle[1:(I-j+1),j-1]))
}

Could you tell more what you want to achieve with that code? Maybe desired output or equation? — Maciej
– Maciej, Commented Mar 17, 2014 at 11:28
I believe not, because your equation involves 1:(j+stuff) and you can't vectorize that. — Carl Witthoft
– Carl Witthoft, Commented Mar 17, 2014 at 11:30
I am trying to fill the vector xi with the formula, xi[j]<-(1/(I-j+1))*sum(log(Triangle[1:(I-j+1),j]/Triangle[1:(I-j+1),j-1])) But i was wondering if I could use some kind of ":" to avoid the loop here — user3235255
– user3235255, Commented Mar 17, 2014 at 11:45
Well, there's always the vectorize function. You'd have to run a time trial to see if it saves you anything in this simple case. — Carl Witthoft
– Carl Witthoft, Commented Mar 17, 2014 at 11:56

shadow · Accepted Answer · 2014-03-17 13:48:06Z

2

Here's a vectorized solution. However, this does not seem to improve efficiency.

mat <- matrix(NA, nrow=I, ncol=J)
smat1 <- cbind(FALSE, rbind(lower.tri(matrix(nrow=I-1, ncol=J-1), diag=TRUE)[(I-1):1,], FALSE))
smat2 <- lower.tri(matrix(nrow=I, ncol=J))[I:1,]
mat[smat1] <- Triangle[smat1] / Triangle[smat2]
res <- c(1, colMeans(log(mat), na.rm=TRUE)[-1])

answered Mar 17, 2014 at 13:48

shadow

22.4k5 gold badges67 silver badges80 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Avoiding a loop in R

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related