My task is to write function in lisp which finds maximum of a list given as argument of the function, by using recursion.I've tried but i have some errors.I'm new in Lisp and i am using cusp plugin for eclipse.This is my code:
(defun maximum (l)
(if (eq((length l) 1)) (car l)
(if (> (car l) (max(cdr l)))
(car l)
(max (cdr l))
))
If this isn't a homework question, you should prefer something like this:
(defun maximum (list)
(loop for element in list maximizing element))
Or even:
(defun maximum (list)
(reduce #'max list))
(Both behave differently for empty lists, though)
If you really need a recursive solution, you should try to make your function more efficient, and/or tail recursive. Take a look at Diego's and Vatine's answers for a much more idiomatic and efficient recursive implementation.
Now, about your code:
It's pretty wrong on the "Lisp side", even though you seem to have an idea as to how to solve the problem at hand. I doubt that you spent much time trying to learn lisp fundamentals. The parentheses are messed up -- There is a closing parenthesis missing, and in ((length l) 1), you should note that the first element in an evaluated list will be used as an operator. Also, you do not really recurse, because you're trying to call max (not maximize). Finally, don't use #'eq for numeric comparison. Also, your code will be much more readable (not only for others), if you format and indent it in the conventional way.
You really should consider spending some time with a basic Lisp tutorial, since your question clearly shows lack of understanding even the most basic things about Lisp, like the evaluation rules.
I see no answers truly recursive and I've written one just to practice Common-Lisp (currently learning). The previous answer that included a recursive version was inefficient, as it calls twice maximum recursively. You can write something like this:
(defun my-max (lst)
(labels ((rec-max (lst actual-max)
(if (null lst)
actual-max
(let ((new-max (if (> (car lst) actual-max) (car lst) actual-max)))
(rec-max (cdr lst) new-max)))))
(when lst (rec-max (cdr lst) (car lst)))))
This is (tail) recursive and O(n).
max_, maxx, ll, m or lst are really not idiomatic. Most lispers strongly prefer hyphenated, self-explaining names. Also, since CL is a Lisp-2, you can bind variables to names like list without problems. Finally, although implementations will usually optimize tail calls (with the right settings), this is not required by the standard. I know this was just an exercise, but iterative/applicative style is often preferred in CL.
lst list, but that's really just nitpicking. Maybe note that (when (not (null lst)) ...) is just (when lst ...) in CL. (In fact, you could even use (and lst (rec-max (cdr lst) (car lst))), but I think I'd prefer when in this case. Seems clearer to me.)I think your problem lies in the fact that you refer to max instead of maximum, which is the actual function name.
This code behaves correctly:
(defun maximum (l)
(if (= (length l) 1)
(car l)
(if (> (car l) (maximum (cdr l)))
(car l)
(maximum (cdr l)))))
maximum twice, which is highly inefficient.As written, that code implies some interesting inefficiencies (it doesn't have them, because you're calling cl:max instead of recursively calling your own function).
Function calls in Common Lisp are typically not memoized, so if you're calling your maximum on a long list, you'll end up with exponential run-time.
There are a few things you can do, to improve the performance.
The first thing is to carry the maximum with you, down the recursion, relying on having it returned to you.
The second is to never use the idiom (= (length list) 1). That is O(n) in list-length, but equivalent to (null (cdr list)) in the case of true lists and the latter is O(1).
The third is to use local variables. In Common Lisp, they're typically introduced by let. If you'd done something like:
(let ((tail-max (maximum (cdr l))))
(if (> (car l) tail-max)
(car l)
tail-max))
You would've had instantly gone from exponential to, I believe, quadratic. If in combination had done the (null (cdr l)) thing, you would've dropped to O(n). If you also had carried the max-seen-so-far down the list, you would have dropped to O(n) time and O(1) space.
if i need to do the max code in iteration not recursive how the code will be ?? i first did an array
(do do-array (d l)
setf b (make-array (length d))
(do (((i=0)(temp d))
((> i (- l 1)) (return))
(setf (aref b i) (car temp))
(setq i (+ i 1))
(setq temp (cdr temp))))
I made this, hope it helps and it is recursive.
(defun compara ( n lista)
(if(endp lista)
n
(if(< n (first lista))
nil
(compara n (rest lista)))))
(defun max_lista(lista)
(if (endp lista)
nil
(if(compara (first lista) (rest lista))
(first lista)
(max_lista(rest lista)))))
A proper tail-recursive solution
(defun maximum (lst)
(if (null lst)
nil
(maximum-aux (car lst) (cdr lst))))
(defun maximum-aux (m lst)
(cond
((null lst) m)
((>= m (car lst)) (maximum-aux m (cdr lst)))
(t (maximum-aux (car lst) (cdr lst)))))
(defun maxx (l)
(if (null l)
0
(if(> (car l) (maxx(cdr l)))
(car l)
(maxx (cdr l)))))
max!=maximum. Also, the function should be written to be tail-call optimized to avoid exploding the stack. Do this by passing all the information up (e.g. max-to-date).maxdo? :)