Loading ...
Sorry, an error occurred while loading the content.

Re: [aima-talk] Complex returns from recursive functions - tail recursion?

Expand Messages
  • Steven Shaw
    ... I can say that it s not tail recursive. I couldn t spot it at first because the simple form to look for is a recursive call embedded in an argument
    Message 1 of 5 , Jan 29, 2006
    • 0 Attachment
      On 29/01/06, Robert Futrelle <futrelle@...> wrote:
      > This is more a straight Lisp question, but it relates to material in
      > the early chapters of PAIP. (Someone can direct me to a better list
      > for this, I'm sure.)
      >
      > This function splits an even-length list into its even and odd
      > (sequence) elements:
      >
      > (defun split-list (lst)
      > (cond ((null lst) '(() ()))
      > (t (let ((lst2 (split-list (rest (rest lst)))))
      > (list (cons (first lst) (first lst2))
      > (cons (second lst) (second lst2)))
      > )
      > )
      > )
      > )
      >
      > Point being, there's more to recursion than consing up a single list.
      >
      > Execution:
      >
      > 1 > (split-list '(b d i o g g))
      > ((B I G) (D O G))
      >
      > I assume it's not tail-recursive.

      I can say that it's not tail recursive. I couldn't spot it at first
      because the simple form to look for is a recursive call embedded in an
      argument position (I don't know the correct terminology), like "(cons
      lst (split-list ...". However, the problem is that the result of the
      recursive call is bound to lst2 and then there is more to do. It's
      just that the lst2 is used twice so it's an optimisation.

      > Can it be made so using multiple-value return?

      Not sure but I think not. Multiple return value is like an
      optimisation for returning a tuple (a pair in this case). Using
      mutiple return values might prevent consing of the pair and use of
      registers for both return values. It doesn't mean change the fact that
      this technique needs to do something (cons) with the results of a call
      to split-list. What do you think?

      > (this is not a quiz, just a question ;-)

      Seems like a quiz :)

      Steve.
      --
      Steven Shaw http://abstractsimplicity.co.uk/
    • Steven Shaw
      Here s version that s tail recursive though I don t think very efficient because of my use of append and list perhaps were not necessary. (defun split-list2
      Message 2 of 5 , Jan 29, 2006
      • 0 Attachment
        Here's version that's tail recursive though I don't think very
        efficient because of my use of append and list perhaps were not
        necessary.

        (defun split-list2 (list)
        (let ((a (list)) (b (list)))
        (split-list2-helper list a b)))

        (defun split-list2-helper (list firsts seconds)
        (cond ((null list) (list firsts seconds))
        (t (split-list2-helper
        (rest (rest list))
        (append firsts (list (first list)))
        (append seconds (list (second list)))))))

        Also, I forget the syntax for defining inner auxilary functions...
        flet, labels, defun??...

        Steve.
      • Steven Shaw
        ... Looks good. I could have done with nvreverse myself!
        Message 3 of 5 , Jan 29, 2006
        • 0 Attachment
          On 29/01/06, Peter Norvig <peter@...> wrote:
          > How about this:

          Looks good. I could have done with nvreverse myself!
        Your message has been successfully submitted and would be delivered to recipients shortly.