s c h e m a t i c s : c o o k b o o k

This Web

Other Webs

Schematics

RSS

Scheme Links

Introduction

Vectors are efficient data structures, containing a fixed number of elements that can be accessed in constant time, and in any order. Vectors are mutable, though 'growing' a vector can be costly (as opposed to lists, which can grow at very little cost).

Vectors are printed in Scheme as a hash mark (#) followed by individual elements contained with a set of parenthesis.

(define test (vector 1 #\a "dog" 0.123 (/ 2 3)))

Like lists, vectors can contain any mixture of types. This ability allows vectors to be used to emulate object-oriented class-like mechanisms by storing functions as elements of the vector:

> (define a (vector 1 'a "dog" 0.123 (lambda (x) (+ 1 2))))
> a
#(1 'a "dog" 0.123 (lambda (x) (+ 1 2)))
> (vector-ref a 2)
"dog"
> ((eval (vector-ref a 4)) 1)
3 

The recipes in this section show various techniques to make the most of the vector data type.

-- BrentAFulgham - 21 Jan 2007

Notes

Although #(1 2 3) evaluates to a vector in PLT Scheme, it is often better to use (vector 1 2 3). The syntax #(1 2 3) indicates a literal constant, so it is an error to mutate the vector. With (vector 1 2 3) the call to vector allocates a new vector, which it is ok to mutate. Note also that in R5RS you need to quote vector constants, that is you need to write '#(1 2 3). Finally normal vectors aren't "growable", but see for example the package "evector.plt" at PLaneT, if you need a growable vector like data structure.

-- JensAxelSoegaard - 21 Jan 2007