CODEWITHSUNDEEP
rvectorizationapplylapplysapply
GeekCat
GeekCat

Reputation: 409

How to iteratively update an variable with its previous value in R with vectorization?

I have a piece of R code like this

s<-vector()
s[1]=1
for(i in 2:10){
   ds= 1/i/s[i-1]
   s[i]= s[i-1]+ds
 }
 s

The result is:

 [1] 1.000000 1.500000 1.722222 1.867384 1.974485 2.058895 2.128281 2.187014
 [9] 2.237819 2.282505

The above is an toy example, what I am trying to do is updating s[i] with some sort of use of s[i-1], please help me with this case, I have tried the following:

s<-vector()
s[1]<-1
sapply(2:10,FUN = function(i){
   if(i==1){
      return(s[i])
   }
   ds<-1/i/s[i-1]
   #print(paste(i,s_new,ds,ds+s[i-1]))
   s[i]= s[i-1]+ds
   return(s[i])
   }
)

The above does not work

Upvotes: 1

Views: 276

Answers (1)

TARehman
TARehman

Reputation: 6749

I think that using <<- will let you directly assign to the global version of s. This is not something I do frequently. See this question for good details of how environments work in R.

I'm not crazy with the below version, mostly because it's mixing a return() with a direct assignment <<-. At this point, it seems like you might as well use a loop as mentioned in the comments.

s <- vector()
s[1] <- 1

sapply(2:10,FUN = function(i){

    if(i == 1){

        return(s[i])
    }
    ds <- 1/i/s[i - 1]
    s[i] <<- s[i - 1] + ds
    }
)

Upvotes: 2

Related Questions