Converting a matrix into an array

Modeling
1

Hi all,
I would like to convert a matrix into an array. Following this How to convert a matrix to an array of vectors?, I wrote the code as:

transformed parameters{
   matrix[J,K] theta_raw;
   real theta[J,K];

theta_raw  = exp(a +b);  // a and b are J by K dimensional matrices 
   for(j in 1:J){
        theta[j,] = theta_raw[j,];    
      }
}

I am getting the following error, which I do not understand why? Please help

SYNTAX ERROR, MESSAGE(S) FROM PARSER:

Dimension mismatch in assignment; variable name = theta, num dimensions given = 1; right-hand side dimensions = 0
Illegal statement beginning with non-void expression parsed as
  theta[j,  ]
Not a legal assignment, sampling, or function statement.  Note that
  * Assignment statements only allow variables (with optional indexes) on the left;
  * Sampling statements allow arbitrary value-denoting expressions on the left.
  * Functions used as statements must be declared to have void returns
2

This code may be valid in R, but not in Stan. I think you can write theta[j,:] = theta_raw[j,:], but I’m not sure.

3

You can write theta[j] = theta_raw[j] to put the jth row of theta_raw into the jth row of theta.

4

Thanks @jjramsey, @hhau.
None of these suggestions worked okay. To do this I had to loop over K as well.

for(j in 1:J){
         for(k in 1:K){
            theta[j,k] = theta_raw[j,k];   
         }
       }

This slows down the code. So thought there must be a way to improve the run time.
Any other suggestions?
Thanks

5

Does to_array_2d not do what you need?:

real theta [J,K] = to_array_2d(theta_raw);
6

Well, you have two parameters, theta and theta_raw, that appear to be exact copies of each other aside from them being slightly different variable types. Just the act of copying itself will bog things down, no matter how efficiently you do it, so I’d suggest avoiding that copying altogether.