[R] fill binary matrices with random 1s

[Bert Gunter]

Folks:

On Tue, Nov 29, 2011 at 6:24 AM, Sarah Goslee <sarah.goslee at gmail.com> wrote:
> I have to admit I'm not entirely sure what your question is. How to
> put a 1 in a random position in a matrix?
>
> mat <- matrix(0, 10, 10)
>  mat[sample(1:nrow(mat), 1), sample(1:ncol(mat), 1)] <- 1
This is unnecessary. In R: matrices are simply vectors with a "dim"
attribute, so you can treat them as vectors:

mat[sample(nrow(mat)*ncol(mat),1] <- 1

Moreover, this also suggests a simple way to do this sequentially:
Simply create your vector of random indices at one go and use it for
your loop -- no checking on what previously was sampled is necessary:

ransamp <- sample(100,100)  ## assuming nrow = ncol = 10

for( i in 1:100) {
  mat[ransamp[i]] <- 1
## do whatever you want
}

HTH

Cheers,
Bert

> will do so, but if you need to fill a random position that is *currently zero*
> then you'll need to wrap it in a while loop and check the value of that cell.
>
> Or, more elegantly, create a random vector of positions in advance,
> then fill each:
> tofill <- sample(1:100)
> for(i in 1:length(tofill)) {
> mat[tofill[i]] <- 1
> }
>
> But if you don't need sequential matrices, just random matrices of
> particular densities, there are nicer ways to create them.
>
> matdensity <- 45
> matsize <- 10
> mat45 <- matrix(sample(c(rep(1, matdensity), rep(0, matsize*2 -
> matdensity))), matsize, matsize)
>
>
> On Tue, Nov 29, 2011 at 7:32 AM, Grant McDonald
> <grantforaccount at hotmail.co.uk> wrote:
>>
>> Dear all, I am finding difficulty in the following, I would like to
>> create an empty matrix e.g. 10x10 of 0s and sequentially fill this
>> matrix with randomly placed a 1s until it is saturated. Producing 100
>> matrices of sequentially increasing density., This process needs to be
>> randomized 1000 times., I assume i should run this along the following
>> lines, 1) Create 1000 matrices all zeros, 2) add a random 1 to all
>> matrices, 3) run function on all 1000 matrices and output results to a
>> vector table (i.e. calculate density of matric at each step for all 100 matrices.
>> )., 4) add another 1 to the previous 1000 matrices in a
>> random position., repeat till all matrices saturated., I have looked
>> through histories on random fill algorithms but all packages I can find
>> nothing as simple as the random fill I am looking for., sorry for
>> bothering, Thank you for any help in advance.
>>
>>
>> Something that starts along the lines of the following? Sorry this example is atrocious.
>>
>> matrixfill <- function(emptymatrix, K=fullmatrix, time=100, from=0, to=time)
>>
>> {
>>
>> N <- numeric(time+1)
>>
>> N[1] <- emptymatrix
>>
>> for (i in 1:time) N[i+1] <- N[i]+"place random 1 in a random xy position" until K.
>> Calculate Density of matrix
>
>
>
> --
> Sarah Goslee
> http://www.functionaldiversity.org
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>



-- 

Bert Gunter
Genentech Nonclinical Biostatistics

Internal Contact Info:
Phone: 467-7374
Website:
http://pharmadevelopment.roche.com/index/pdb/pdb-functional-groups/pdb-biostatistics/pdb-ncb-home.htm