Question

I have a simple R code to add up the components of a 2×2 matrix

sum<-0
for(i in 1:2){ # row
    for(j in 1:2){ #column
      sum<-sum+mat[i,j]
    }
  }

Is it possible to use the outer(), or any other function to replace this piece of code and make it more efficient? My aim is to replace nested for loops in my entire code to bring down the time needed to execute my program.

Edit:
I also want to try using it on code snippets like:

for(i in 1:2){ # row
    for(j in 1:2){ #coloumn

      chisqr<- chisqr+ ((mat[i,j]-expmat[i,j])^2)/expmat[i,j]

    }
  }

and:

  for(i in 1:2){ # row
    for(j in 1:2){ #coloumn
      rowsum<-0
      colsum<-0

      for(k in 1:2){
        rowsum<- rowsum+mat[i,k]
      }

      for(k in 1:2){
        colsum<- colsum+mat[k,j]
      }

      expmat[i,j]<- (rowsum*colsum)/sum
    }
  }

Answer 1

sum(mat)

will do the trick. There’s no need for a loop or outer.

---

Update based on new questions:

You could calculate expmat using outer:

outer(rowSums(mat), colSums(mat))/sum(mat)

---

Afterwards, you could calculate chisqr:

sum((mat-expmat)^2/expmat)

By the way: I recommend having a look at ?chisq.test.