Why operator function not work?

Modeling
#1

I put a function if_greater in the functions block like this:

 functions {
     vector if_greater(vector x, vector y) {		     
         int n;		        
	vector[n] D;
        n = rows(x);		         
         for (i in 1:n){		       
            D[i] = operator>(x[i],y[i]) ;		     
         }		        
          return D;		
     }

An error popped up and pointed to the operator> function. I changed it to the conditional operator function x[i] > y[i] ? 1:0 and it works!

I looked up the reference manual. the operator function can be used in this way. So I wonder why the error occurred. Thanks.

#2

The entire function would be better written as

functions {
  vector if_greater(vector x, vector y) {
    int n = rows(x);
    vector[n] D;
    if (rows(y) != n) reject("rows of x and y must match");
    for (i in 1:n) D[i] = x[i] > y[i];
    return D;
  }
}
#3

I think there’s some confusion here about what operator means, which is totally understandable if you’re not used to that terminology. operator isn’t the name of a function, it’s a term used to describe the type of construct that > is, which is basically a function that behaves differently semantically than regular functions (https://en.wikipedia.org/wiki/Operator_(computer_programming)). In other words, describing > as an operator is intended to convey that the syntax to use it is x > y rather than >(x,y), which would be the standard way to call regular functions. Anyway, you can just use @bgoodri’s version of the function, but to get your original version to not error you’d need to replace operator>(x[i],y[i]) with just x[i] > y[i]. Hope that helps clarify.

1 Like
#4

Thanks! It’s truly helpful!

#5

Thanks for clarification!

1 Like