In an answer which explains how to convert a number to a list the number->list procedure is defined as follows:
(define (number->list n)
(let loop ((n n)
(acc '()))
(if (< n 10)
(cons n acc)
(loop (quotient n 10)
(cons (remainder n 10) acc)))))
Here a "named let" is used. I don't understand how this named let works.
I see that a loop is defined where the variable n is equal to n, and the variable acc equal to the empty list. Then if n is smaller than 10 the n is consed to the acc. Otherwise, "the loop" is applied with n equal to n/10 and acc equal to the cons of the remainder of n/10 and the previous accumulated stuff, and then calls itself.
I don't understand why loop is called loop (what is looping?), how it can automatically execute and call itself, and how it will actually add each number multiplied by its appropriate multiplier to form a number in base 10.
I hope someone can shine his or her light on the procedure and the above questions so I can better understand it. Thanks.
The basic idea behind a named let is that it allows you to create an internal function, that can call itself, and invoke it automatically. So your code is equivalent to:
(define (number->list n)
(define (loop n acc)
(if (< n 10)
(cons n acc)
(loop (quotient n 10)
(cons (remainder n 10) acc))))
(loop n '()))
Hopefully, that is easier for you to read and understand.
You might, then, ask why people tend to use a named let rather than defining an internal function and invoking it. It's the same rationale people have for using (unnamed) let: it turns a two-step process (define a function and invoke it) into one single, convenient form.
It's called a loop because the function calls itself in tail position. This is known as tail recursion. With tail recursion, the recursive call returns directly to your caller, so there's no need to keep the current call frame around. You can do tail recursion as many times as you like without causing a stack overflow. In that way, it works exactly like a loop.
If you'd like more information about named let and how it works, I wrote a blog post about it. (You don't need to read it to understand this answer, though. It's just there if you're curious.)
A normal let usage can be considered an anonymous procedure call:
(let ((a 10) (b 20))
(+ a b))
;; is the same as
((lambda (a b)
(+ a b))
10
20)
A named let just binds that procedure to a name in the scope of the procedure so that it is equal to a single procedure letrec:
(let my-chosen-name ((n 10) (acc '()))
(if (zero? n)
acc
(my-chosen-name (- n 1) (cons n acc)))) ; ==> (1 2 3 4 5 6 7 8 9 10)
;; Is the same as:
((letrec ((my-chosen-name
(lambda (n acc)
(if (zero? n)
acc
(my-chosen-name (- n 1) (cons n acc))))))
my-chosen-name)
10
'()) ; ==> (1 2 3 4 5 6 7 8 9 10)
Notice that the body of the letrec just evaluates to the named procedure so that the name isn't in the environment of the first call. Thus you could do this:
(let ((loop 10))
(let loop ((n loop))
(if (zero? n)
'()
(cons n (loop (- n 1))))) ; ==> (10 9 8 7 6 5 4 3 2 1)
the procedure loop is only in the environment of the body of the inner let and does not shadow the variable loop of the outer let.
In your example, the name loop is just a name. In Scheme every loop is ultimately done with recursion, but usually the name is used when it's tail recursion and thus an iterative process.
Related
I want to find a way to use a counter within a function where the counter is also a variable within the recursive function. An example of this is with a program that takes a list and then "sifts" through it until it finds all the numbers within it that are multiples of i:
(define (multiples-of lst) (lambda (i) (if (> i 3))
'()
(multiplefilter (ismultipleof (+ i 1)) (lst)))))
where ismultipleof checks if the car of each list is a multiple of i + 1 (with i starting at 1) and then the multiplefilter is a separate function that scraps any values of the list that are not multiples of i. So that if I put in the list (1 2 3 4 5 6 7 8 9 10 11 12) the output would just be (6 12). The biggest issue is getting said counter to work within the function.
Well, here's a simple function that uses a counter in its execution. range just returns a list of integers from 0 up to (- n 1). It's just as easy to write something that increments a counter.
(define (range n)
(let loop ((acc '())
(list-start (- n 1)))
(if (negative? list-start)
acc
(loop (cons list-start acc) (- list-start 1)))))
This function counts down rather than up because recursion usually builds lists "backwards". Counting down just avoids the need call reverse on the acc when the recursion is finished.
Iteration is usually accomplished with recursion in Scheme instead of for loops as in some languages. In particular, recursion from the tail is something to strive for since it will not "blow the stack".
I'm running a for loop in racket, for each object in my list, I want to execute two things: if the item satisfies the condition, (1) append it to my new list and (2) then print the list. But I'm not sure how to do this in Racket.
This is my divisor function: in the if statement, I check if the number in the range can divide N. If so, I append the item into my new list L. After all the loops are done, I print L. But for some unknown reason, the function returns L still as an empty list, so I would like to see what the for loop does in each loop. But obviously racket doesn't seem to take two actions in one "for loop". So How should I do this?
(define (divisor N)
(define L '())
(for ([i (in-range 1 N)])
(if (equal? (modulo N i) 0)
(append L (list i))
L)
)
write L)
Thanks a lot in advance!
Note: This answer builds on the answer from #uselpa, which I upvoted.
The for forms have an optional #:when clause. Using for/fold:
#lang racket
(define (divisors N)
(reverse (for/fold ([xs '()])
([n (in-range 1 N)]
#:when (zero? (modulo N n)))
(displayln n)
(cons n xs))))
(require rackunit)
(check-equal? (divisors 100)
'(1 2 4 5 10 20 25 50))
I realize your core question was about how to display each intermediate list. However, if you didn't need to do that, it would be even simpler to use for/list:
(define (divisors N)
(for/list ([n (in-range 1 N)]
#:when (zero? (modulo N n)))
n))
In other words a traditional Scheme (filter __ (map __)) or filter-map can also be expressed in Racket as for/list using a #:when clause.
There are many ways to express this. I think what all our answers have in common is that you probably want to avoid using for and set! to build the result list. Doing so isn't idiomatic Scheme or Racket.
It's possible to create the list as you intend, as long as you set! the value returned by append (remember: append does not modify the list in-place, it creates a new list that must be stored somewhere) and actually call write at the end:
(define (divisor N)
(define L '())
(for ([i (in-range 1 N)]) ; generate a stream in the range 1..N
(when (zero? (modulo N i)) ; use `when` if there's no `else` part, and `zero?`
; instead of `(equal? x 0)`
(set! L (append L (list i))) ; `set!` for updating the result
(write L) ; call `write` like this
(newline)))) ; insert new line
… But that's not the idiomatic way to do things in Scheme in general and Racket in particular:
We avoid mutation operations like set! as much as possible
It's a bad idea to write inside a loop, you'll get a lot of text printed in the console
It's not recommended to append elements at the end of a list, that'll take quadratic time. We prefer to use cons to add new elements at the head of a list, and if necessary reverse the list at the end
With all of the above considerations in place, this is how we'd implement a more idiomatic solution in Racket - using stream-filter, without printing and without using for loops:
(define (divisor N)
(stream->list ; convert stream into a list
(stream-filter ; filter values
(lambda (i) (zero? (modulo N i))) ; that meet a given condition
(in-range 1 N)))) ; generate a stream in the range 1..N
Yet another option (similar to #uselpa's answer), this time using for/fold for iteration and for accumulating the value - again, without printing:
(define (divisor N)
(reverse ; reverse the result at the end
(for/fold ([acc '()]) ; `for/fold` to traverse input and build output in `acc`
([i (in-range 1 N)]) ; a stream in the range 1..N
(if (zero? (modulo N i)) ; if the condition holds
(cons i acc) ; add element at the head of accumulator
acc)))) ; otherwise leave accumulator alone
Anyway, if printing all the steps in-between is necessary, this is one way to do it - but less efficient than the previous versions:
(define (divisor N)
(reverse
(for/fold ([acc '()])
([i (in-range 1 N)])
(if (zero? (modulo N i))
(let ([ans (cons i acc)])
; inefficient, but prints results in ascending order
; also, `(displayln x)` is shorter than `(write x) (newline)`
(displayln (reverse ans))
ans)
acc))))
#sepp2k has answered your question why your result is always null.
In Racket, a more idiomatic way would be to use for/fold:
(define (divisor N)
(for/fold ((res null)) ((i (in-range 1 N)))
(if (zero? (modulo N i))
(let ((newres (append res (list i))))
(displayln newres)
newres)
res)))
Testing:
> (divisor 100)
(1)
(1 2)
(1 2 4)
(1 2 4 5)
(1 2 4 5 10)
(1 2 4 5 10 20)
(1 2 4 5 10 20 25)
(1 2 4 5 10 20 25 50)
'(1 2 4 5 10 20 25 50)
Since append is not really performant, you usually use cons instead, and end up with a list you need to reverse:
(define (divisor N)
(reverse
(for/fold ((res null)) ((i (in-range 1 N)))
(if (zero? (modulo N i))
(let ((newres (cons i res)))
(displayln newres)
newres)
res))))
> (divisor 100)
(1)
(2 1)
(4 2 1)
(5 4 2 1)
(10 5 4 2 1)
(20 10 5 4 2 1)
(25 20 10 5 4 2 1)
(50 25 20 10 5 4 2 1)
'(1 2 4 5 10 20 25 50)
To do multiple things in a for-loop in Racket, you just write them after each other. So to display L after each iteration, you'd do this:
(define (divisor N)
(define L '())
(for ([i (in-range 1 N)])
(if (equal? (modulo N i) 0)
(append L (list i))
L)
(write L))
L)
Note that you need parentheses to call a function, so it's (write L) - not write L. I also replaced your write L outside of the for-loop with just L because you (presumably) want to return L from the function at the end - not print it (and since you didn't have parentheses around it, that's what it was doing anyway).
What this will show you is that the value of L is () all the time. The reason for that is that you never change L. What append does is to return a new list - it does not affect the value of any its arguments. So using it without using its return value does not do anything useful.
If you wanted to make your for-loop work, you'd need to use set! to actually change the value of L. However it would be much more idiomatic to avoid mutation and instead solve this using filter or recursion.
'Named let', a general method, can be used here:
(define (divisors N)
(let loop ((n 1) ; start with 1
(ol '())) ; initial outlist is empty;
(if (< n N)
(if(= 0 (modulo N n))
(loop (add1 n) (cons n ol)) ; add n to list
(loop (add1 n) ol) ; next loop without adding n
)
(reverse ol))))
Reverse before output since items have been added to the head of list (with cons).
In an effort to find a simple example of CPS which doesn't give me a headache , I came across this Scheme code (Hand typed, so parens may not match) :
(define fact-cps
(lambda(n k)
(cond
((zero? n) (k 1))
(else
(fact-cps (- n 1)
(lambda(v)
(k (* v n))))))))
(define fact
(lambda(n)
(fact-cps n (lambda(v)v)))) ;; (for giggles try (lambda(v)(* v 2)))
(fact 5) => 120
Great, but Scheme isn't Common Lisp, so I took a shot at it:
(defun not-factorial-cps(n k v)
(declare (notinline not-factorial-cps)) ;; needed in clisp to show the trace
(cond
((zerop n) (k v))
((not-factorial-cps (1- n) ((lambda()(setq v (k (* v n))))) v))))
;; so not that simple...
(defun factorial(n)
(not-factorial-cps n (lambda(v)v) 1))
(setf (symbol-function 'k) (lambda(v)v))
(factorial 5) => 120
As you can see, I'm having some problems, so although this works, this has to be wrong. I think all I've accomplished is a convoluted way to do accumulator passing style. So other than going back to the drawing board with this, I had some questions: Where exactly in the Scheme example is the initial value for v coming from? Is it required that lambda expressions only be used? Wouldn't a named function accomplish more since you could maintain the state of each continuation in a data structure which can be manipulated as needed? Is there in particular style/way of continuation passing style in Common Lisp with or without all the macros? Thanks.
The problem with your code is that you call the anonymous function when recurring instead of passing the continuation like in the Scheme example. The Scheme code can easily be made into Common Lisp:
(defun fact-cps (n &optional (k #'values))
(if (zerop n)
(funcall k 1)
(fact-cps (- n 1)
(lambda (v)
(funcall k (* v n))))))
(fact-cps 10) ; ==> 3628800
Since the code didn't use several terms or the implicit progn i switched to if since I think it's slightly more readable. Other than that and the use of funcall because of the LISP-2 nature of Common Lisp it's the identical code to your Scheme version.
Here's an example of something you cannot do tail recursively without either mutation or CPS:
(defun fmapcar (fun lst &optional (k #'values))
(if (not lst)
(funcall k lst)
(let ((r (funcall fun (car lst))))
(fmapcar fun
(cdr lst)
(lambda (x)
(funcall k (cons r x)))))))
(fmapcar #'fact-cps '(0 1 2 3 4 5)) ; ==> (1 1 2 6 24 120)
EDIT
Where exactly in the Scheme example is the initial value for v coming
from?
For every recursion the function makes a function that calls the previous continuation with the value from this iteration with the value from the next iteration, which comes as an argument v. In my fmapcar if you do (fmapcar #'list '(1 2 3)) it turns into
;; base case calls the stacked lambdas with NIL as argument
((lambda (x) ; third iteration
((lambda (x) ; second iteration
((lambda (x) ; first iteration
(values (cons (list 1) x)))
(cons (list 2) x)))
(cons (list 3) x))
NIL)
Now, in the first iteration the continuation is values and we wrap that in a lambda together with consing the first element with the tail that is not computed yet. The next iteration we make another lambda where we call the previous continuation with this iterations consing with the tail that is not computed yet.. At the end we call this function with the empty list and it calls all the nested functions from end to the beginning making the resulting list in the correct order even though the iterations were in oposite order from how you cons a list together.
Is it required that lambda expressions only be used? Wouldn't a named
function accomplish more since you could maintain the state of each
continuation in a data structure which can be manipulated as needed?
I use a named function (values) to start it off, however every iteration of fact-cps has it's own free variable n and k which is unique for that iteration. That is the data structure used and for it to be a named function you'd need to use flet or labels in the very same scope as the anonymous lambda functions are made. Since you are applying previous continuation in your new closure you need to build a new one every time.
Is there in particular style/way of continuation passing style in
Common Lisp with or without all the macros?
It's the same except for the dual namespace. You need to either funcall or apply. Other than that you do it as in any other language.
I am new to scheme, and have the following question:
If I want a function to also print -the value- of an expression and then call a function, how would one come up to doing that?
For example, I need the function foo(n) to print the value of n mod 2 and call foo(n/2), I would've done:
(define foo (lambda (n) (modulo n 2) (foo (/ n 2))))
But that, of course, would not print the value of n mod 2.
Here is something simple:
(define foo
(lambda (n)
(display (modulo n 2))
(when (positive? n)
(foo (/ n 2)))))
Note the check of (positive? n) to ensure that you avoid (/ 0 2) forever and ever.
I'm terrible at Lisp, but here's an idea: Maybe you could define a function that prints a value and returns it
(define (debug x) (begin (display x) (newline) x))
Then just call the function like
(some-fun (debug (some expression)))
As #Juho wrote, you need to add a display. But, your procedure is recursive without a base case, so it will never terminate.
Try this:
(define foo
(lambda (n)
(cond
((integer? n) (display (modulo n 2))
(newline)
(foo (/ n 2)))
(else n))))
then
> (foo 120)
0
0
0
1
7 1/2
Usually when dealing with more than one thing it's common to build lists to present a solution when the procedure is finished.
(define (get-digits number base)
(let loop ((nums '()) (cur number))
(if (zero? cur)
nums
(loop (cons (remainder cur base) nums)
(quotient cur base)))))
(get-digits 1234 10) ; ==> (1 2 3 4)
Now, since you use DrRacket you have a debugger so you can actually step though this code but you rather should try to make simple bits like this that is testable and that does not do side effects.
I was puzzled when you were taling about pink and blue output until I opened DrRacket and indeed there it was. Everything that is pink is from the program and everything blue is normally not outputed but since it's the result of top level forms in the IDE the REPL shows it anyway. The differences between them are really that you should not rely on blue output in production code.
As other has suggested you can have debug output with display within the code. I want to show you another way. Imagine I didn't know what to do with the elements so I give you the opportunity to do it yourself:
(define (get-digits number base glue)
(let loop ((nums '()) (cur number))
(if (zero? cur)
nums
(loop (glue (remainder cur base) nums)
(quotient cur base)))))
(get-digits 1234 10 cons) ; ==> (1 2 3 4)
(define (debug-glue a d)
(display a)
(newline)
(cons a d))
(get-digits 1234 10 debug-glue) ; ==> (1 2 3 4) and displays "4\n3\n2\n1\n"
Does scheme have a function to call a function n times. I don't want map/for-each as the function doesn't have any arguments. Something along the lines of this :-
(define (call-n-times proc n)
(if (= 0 n)
'()
(cons (proc) (call-n-times proc (- n 1)))))
(call-n-times read 10)
SRFI 1 has a list-tabulate function that can build a list from calling a given function, with arguments 0 through (- n 1). However, it does not guarantee the order of execution (in fact, many implementations start from (- n 1) and go down), so it's not ideal for calling read with.
In Racket, you can do this:
(for/list ((i 10))
(read))
to call read 10 times and collect the result of each; and it would be done left-to-right. But since you tagged your question for Guile, we need to do something different.
Luckily, Guile has SRFI 42, which enables you to do:
(list-ec (: i 10)
(read))
Implementing tail-recursion modulo cons optimization by hand, to build the resulting list with O(1) extra space:
(define (iterate0-n proc n) ; iterate a 0-arguments procedure n times
(let ((res (list 1))) ; return a list of results in order
(let loop ((i n) (p res))
(if (< i 1)
(cdr res)
(begin
(set-cdr! p (list (proc)))
(loop (- i 1) (cdr p)))))))
This technique first (?) described in Friedman and Wise's TR19.