You example works. Thank you. Two more question:
1. Can we declare a float (single or double) array output? When I changed "int32" into "dbl", the output "y" also turned into DBL. But it was not an array anymore.
2. In your example, there is a 'for' loop to assign a value to each component in array "y". Is this 'for' loop necessary? I mean, can we do it like that polymorphic way?