Reassign values within a Loop in R

Question

I need to multiply the matrix estado by the matrix matriz_de transition and re-assign the variable estado to the result of the multiplication, I have to do this 1000 times, and I need to repeat the simulation 100 times.

I.e Multiply the estado matrix by the matriz_de_transicion 1000 times and see if the value of estado converges, then repeat this operation 100 times.

I have tried with for loops, but the way I re-assign the values to the variable estado doesn't work since I keep getting the same results, even if when I do it manually it seems to work.

This is my code:

    nombre_estados <- c("Estado 1","Estado 2","Estado 3")
    matriz_de_transicion <- matrix(c(0.2,0.7,0.1,
                                     0.3,0.7,0.0,
                                     0.1,0.4,0.5), 
                            byrow = T,nrow = 3, dimnames = list(nombre_estados,nombre_estados))
    estado <- matrix(c(0.2,0.7,0.1),byrow = T, nrow = 1)

    estado <- estado %*% matriz_de_transicion
    estado # 0.26     0.67     0.07
    estado <- estado %*% matriz_de_transicion
    estado # 0.26    0.679    0.061
   # repeat this 1000 times

Many thanks for the answers,how could I then add these final estados to the 100 x 3 matrix?. I created the 100 x 3 matrix:

datos <- matrix(c(0,0,0),byrow = F,ncol =  3, nrow = 100)

Then I tried to do a nested loop:

for(i in 1:100){
for(i in 1:1000){
  estado <- estado %*% matriz_de_transicion  
}
  datos[,1] <- estado[1,1]
  datos[,2] <- estado[1,2]
  datos[,3] <- estado[1,3]
  estado <- matrix(c(0.2,0.7,0.1),byrow = T, nrow = 1) # I thought this would reset the value to the initial state but I fill the 100 x 3 matrix with the same values.
}
           [,1]      [,2]       [,3]
  [1,] 0.2631579 0.6842105 0.05263158
  [2,] 0.2631579 0.6842105 0.05263158
  [3,] 0.2631579 0.6842105 0.05263158
  [4,] 0.2631579 0.6842105 0.05263158
  [5,] 0.2631579 0.6842105 0.05263158
  [6,] 0.2631579 0.6842105 0.05263158
  [7,] 0.2631579 0.6842105 0.05263158
  [8,] 0.2631579 0.6842105 0.05263158

I'm struggling to reset the value of the estado variable,for each one of the 100 simulations I need to obtain in the end, so each one of the 1:1000 loops have the same starting value for estado.

Rui Barradas · Accepted Answer · 2018-05-13 10:53:09Z

1

Maybe something like the following.

I have the function return extra information, since you ask whether the value of estado converges.

fun <- function(M, Trans, n = 1000, tol = .Machine$double.eps^0.5){
    for(i in seq_len(n)){
        Prev <- M
        M <- M %*% Trans
        if(all(abs(M - Prev) < tol)) break
    }
    list(Final = M, converge = i < n, iter = i)
}

fun(estado, matriz_de_transicion)
#$`Final`
#      Estado 1  Estado 2   Estado 3
#[1,] 0.2631579 0.6842105 0.05263159
#
#$converge
#[1] TRUE
#
#$iter
#[1] 18

answered May 13, 2018 at 10:53

Rui Barradas

78k8 gold badges41 silver badges75 bronze badges

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

3 Comments

MKR Over a year ago

Our values are matching. Hence, I think, simple for-loop is working to fit the scenario.

Rui Barradas Over a year ago

@MKR Yes, but suppose the OP has to do this with different matrices. Then a function is a more convenient way to do it.

MKR Over a year ago

Definitely a function is a better approach. But I was a bit confused with a OP's comment that for-loop doesn't work the way values is re-assign. Hence, I thought to clarify that with an answer.

MKR · Accepted Answer · 2018-05-13 10:52:52Z

1

I'm sure just a for-loop will fit this purpose. Not sure why OP thinks calculation is not taking into effect for next calls. estado is declared at global scope anyway.

estado <- matrix(c(0.2,0.7,0.1),byrow = T, nrow = 1)
for(i in 1:1000){
  estado <- estado %*% matriz_de_transicion  
}
estado 
#  Estado 1  Estado 2   Estado 3
#[1,] 0.2631579 0.6842105 0.05263158

answered May 13, 2018 at 10:52

MKR

20.2k4 gold badges26 silver badges36 bronze badges

1 Comment

Sergi Over a year ago

Many thanks! How could I then add these final estados to the 100 x 3 matrix?

datos <- matrix(c(0,0,0),byrow = F,ncol =  3, nrow = 100) n = 100 for(i in 1:100){ for(i in 1:1000){   estado <- estado %*% matriz_de_transicion   }   datos[,1] <- estado[1,1]   datos[,2] <- estado[1,2]   datos[,3] <- estado[1,3]   estado <- matrix(c(0.2,0.7,0.1),byrow = T, nrow = 1) }

I'm struggling to reset the value of the estado variable,for each one of the 100 simulations I need to obtain in the end, so each one of the 1:1000 loops have the same starting value for estado.

Collectives™ on Stack Overflow

Reassign values within a Loop in R

2 Answers 2

3 Comments

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

3 Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related