< a quantity that can be divided into another a whole number of time />

Sequences in Lisp

July 16, 2022

In Common Lisp, sequences are made of lists, vectors, strings and [maybe other things]. In short, these are ordered collections of elements (or items for short). You can define a generic sequence using maque-sequence, or rely on specialized functions (make-list, make-array,1 list, vector, etc.) as shown below:

* (defparameter *x* (make-sequence '(vector double-float) 5 :initial-element 1d0))
* *x*
#(1.0d0 1.0d0 1.0d0 1.0d0 1.0d0)
* (vector 1d0 1d0 1d0 1d0 1d0)
#(1.0d0 1.0d0 1.0d0 1.0d0 1.0d0)

Note that although they look the same, make-sequence in this case returned an array, and not a simple vector like vector. Also note that I’m using SBCL REPL so that ** means the last computed result.

* (equalp *x* **)
* (type-of *x*)
(simple-array double-float (5))

To generate a regular sequence à la Python’s range function, we could define the following function:

* (defun iseq (end &optional (start 0) (step 1))
    (loop :for n :from start :below end :by step :collect n))
* (iseq 5)
(0 1 2 3 4)
* (iseq 6 1)
(1 2 3 4 5)

This will only work for numbers, of course. While playing with newLISP, I learned that there’s a builtin sequence function there, which works like LispStat’s iseq:2

> (setq x (sequence 1 5))
(1 2 3 4 5)
> (length x)
> (x 3)

There’s even a series function to create a geometric sequence of numbers. As shown in the above example, subsetting is a breeze (this is called implicit indexing per the documentation), but there’s more: slicing a list or an array is even easier than in most Lisp-like languages I know, except perhaps Clojure using subvec provided you’re using vectors:

> (2 x)
(3 4 5)

Of course, we could do the same with arrays as shown below:

> (setq z (array 2 2 '(1 2 3 4)))
((1 2) (3 4))
> (0 1 z)
((1 2))
> (1 1 z)
((3 4))

I think I’ll need to investigate more into this dialect to see if toy examples from LispStat could be ported to newLISP. Despite all the criticisms we can found on HN, I believe newLISP provides a handy way of writing quick scripts in Lisp, but see I am not afraid to admit that I’ve used Lisp for real work. See also newlisp - an intriguing dialect of Lisp, by Eli Bendersky, which also summarizes the point raised above. Other candidates on my list are PicoLisp and Janet. Stay tuned.

♪ As Animals • By My Side

  1. See also this warning about portable scientific code in CL, especially when dealing with arrays. ↩︎

  2. Or various implementation of the iota function in CL, e.g. Generating sequences in Common Lisp or the Alexandria library. ↩︎

See Also

» Common Lisp Hyperspec in Vim » Generating power sets in Lisp » QR factorization and linear regression » Bootstrap resampling in Lisp » Reading list on Lisp & Co.