How to create 1000 Randomly generated 2x2 Matrices range (-10 and 10) - r

Question

I'm trying to complete a homework problem
I was successful on creating a randomly generated matrix but I do not know how to create more than one in one go.
runif(4,min=-10,max=10)
m=matrix(runif(4*n), ncol = 2, nrow = 2)
no gonna lie I don't really know what I'm doing

Answer 1

You could try:
a<- replicate(100, matrix(runif(4), ncol = 2))
Then you have to access each one as an array which means you have to subset in the third dimension of the array like:
> a[,,2]
[,1] [,2]
[1,] 0.8476489 0.6139453
[2,] 0.1315417 0.8195134
The other way would be with a loop that would generate list objects which are a little easier IMHO to subset.
my_list <- list()
for(i in 1:100){
my_list[[i]]<-matrix(runif(4), ncol = 2)
}
Then you access each element with my_list[[100]]

Answer 2

We can do this using lapply, two main ways, the second is easier to understand, the first way is much more efficient, especially when n gets large,
vals <- runif(n*4, min=-10, max=10)
rst2 <- lapply(1:1000, function(i) matrix(vals[i:(i+3)], ncol = 2))
Or,
the.list <- vector(1000, mode = "list") # Create an empyty list
matrix.list <- lapply(the.list, function(x) matrix(runif(4,min=-10,max=10), nrow = 2))
Returns,
.....
[[999]]
[,1] [,2]
[1,] -6.520801 -5.944080
[2,] -4.183131 1.190629
[[1000]]
[,1] [,2]
[1,] 1.208202 7.86769
[2,] -2.672111 -8.33435
Faster version, first generating all the data needed, then splitting it, then converting the split vectors all into matrices,
vals <- runif(1000*4, min=-10, max=10)
rst <- split(matrices, rep(1:1000, each = 2))
rst2 <- lapply(rst, function(x) matrix(x, ncol = 2))
Fast Loop Using #MDEWITT answer as a base, we can do some simple preprocessing tasks to increase the efficiency.
my_list <- vector(n, mode = "list")
vals <- runif(n*4, min=-10, max=10)
for(i in 1:n){
my_list[[i]]<-matrix(vals[i:(i+3)], ncol = 2,nrow=2)
}
Quick Benchmarking
We see the FLoop and FLapply functions are faster, this becomes more apparent when n=10000....
n <- 1000
microbenchmark(fun1(n), fun2(n), fun3(n), times = 1000)
Unit: milliseconds
expr min lq mean median uq max neval
Lapply(n) 9.709308 9.993209 19.110734 10.708773 12.259961 2515.7737 1000
Predefine(n) 5.159808 5.389392 11.058017 5.615735 6.673107 1004.6575 1000
MDEWITT(n) 10.838335 11.391154 22.738093 12.243137 13.898335 646.6250 1000
FLoop(n) 5.104331 5.384800 9.139668 5.612018 6.502171 369.3693 1000
FLapply(n) 6.191827 6.687455 10.947632 7.613829 8.667169 274.9425 1000
Code used,
library(microbenchmark)
Lapply <- function(n=1000){
the.list <- vector(n, mode = "list")
matrix.list <- lapply(the.list, function(x) matrix(runif(4,min=-10,max=10), nrow = 2))
}
Predefine <- function(n=1000){
vals <- runif(n*4, min=-10, max=10)
rst <- split(vals, rep(1:n, each = 2))
rst2 <- lapply(rst, function(x) matrix(x, ncol = 2))
}
MDEWITT <- function(n=1000){
a <- replicate(n, matrix(runif(4*2), ncol = 2, nrow = 2))
}
FLoop <- function(n=1000){
my_list <- vector(n, mode = "list")
vals <- runif(n*4, min=-10, max=10)
for(i in 1:n){
my_list[[i]]<-matrix(vals[i:(i+3)], ncol = 2,nrow=2)
}
}
FLapply <- function(n=1000){
vals <- runif(n*4, min=-10, max=10)
rst2 <- lapply(1:n, function(i) matrix(vals[i:(i+3)], ncol = 2))
}