Data types

Clearly, in order to do any useful work, we must define some set of data types we wish to work with. Again, the virtual machine does not yet implement everything one could wish for, but it does implement the most basic types. The types are outlined in Table 3.2.1.

Table 3.1: List of data types currently implemented in the virtual machine.

Type	Description
integer	A 32 bit signed integer, used especially as loop variable and for indexing in other types
floating	A 64 bit IEEE floating point type
tuple	A tuple can hold any number of elements. The elements can be of any type. This is mostly used for argument passing, where it is only possible to pass one argument to a function. One can pass a tuple holding any number of elements (arguments), since the tuple is just one argument.
vector	This is a vector that holds elements of type floating.
matrix	This is a matrix holding elements of type floating.

I felt these were the most basic and important types. The tuple type is important because one can only pass one argument to a sub-routine call, and sub-routines can only return one argument. The tuple type can hold any number of elements of any type (even tuples of tuples of ...), thereby allowing us to pass several variables back and forth between sub-routine calls.

Some languages have multi-dimensional types. I have not yet implemented $n$-dimensional arrays, but in a sense the tuple type is already $n$-dimensional, so the basic functionality for handling $n$-dimensional variables is already present in the system. In the meantime one can use a tuple of vectors or matrices, to simulate $n$-dimensional arrays.