结合记忆和尾递归
是否有可能以某种方式将记忆和尾递归结合起来?我目前正在学习 F#,理解这两个概念,但似乎无法将它们结合起来。
假设我有以下 memoize
函数(来自 真实世界函数式编程) :
let memoize f = let cache = new Dictionary<_, _>()
(fun x -> match cache.TryGetValue(x) with
| true, y -> y
| _ -> let v = f(x)
cache.Add(x, v)
v)
以及以下 factorial
函数:
let rec factorial(x) = if (x = 0) then 1 else x * factorial(x - 1)
记忆 factorial
并不太难,使其尾递归也不是太困难:
let rec memoizedFactorial =
memoize (fun x -> if (x = 0) then 1 else x * memoizedFactorial(x - 1))
let tailRecursiveFactorial(x) =
let rec factorialUtil(x, res) = if (x = 0)
then res
else let newRes = x * res
factorialUtil(x - 1, newRes)
factorialUtil(x, 1)
但是你能将记忆化和尾递归结合起来吗?我做了一些尝试,但似乎无法使其发挥作用。或者这根本不可能?
Is it possible to combine memoization and tail-recursion somehow? I'm learning F# at the moment and understand both concepts but can't seem to combine them.
Suppose I have the following memoize
function (from Real-World Functional Programming):
let memoize f = let cache = new Dictionary<_, _>()
(fun x -> match cache.TryGetValue(x) with
| true, y -> y
| _ -> let v = f(x)
cache.Add(x, v)
v)
and the following factorial
function:
let rec factorial(x) = if (x = 0) then 1 else x * factorial(x - 1)
Memoizing factorial
isn't too difficult and making it tail-recursive isn't either:
let rec memoizedFactorial =
memoize (fun x -> if (x = 0) then 1 else x * memoizedFactorial(x - 1))
let tailRecursiveFactorial(x) =
let rec factorialUtil(x, res) = if (x = 0)
then res
else let newRes = x * res
factorialUtil(x - 1, newRes)
factorialUtil(x, 1)
But can you combine memoization and tail-recursion? I made some attempts but can't seem to get it working. Or is this simply not possible?
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论
评论(5)
与往常一样,延续产生了一个优雅的尾部调用解决方案:
有两种测试。首先,这个演示调用 F(4) 缓存 F(4)、F(3)、F(2)、F(1),如您所愿。
然后,注释掉
***
printf 并取消注释最终测试(并在发布模式下编译)以显示它没有 StackOverflow(它正确使用了 tailcalls)。也许我会概括出“memoize”并在“fib”上演示它接下来...
编辑
好吧,我想这是下一步,将记忆与阶乘解耦:
编辑
好吧,这是一个似乎有效的完全概括的版本。
As always, continuations yield an elegant tailcall solution:
There are two kinds of tests. First, this demos that calling F(4) caches F(4), F(3), F(2), F(1) as you would like.
Then, comment out the
***
printf and uncomment the final test (and compile in Release mode) to show that it does not StackOverflow (it uses tailcalls correctly).Perhaps I'll generalize out 'memoize' and demonstrate it on 'fib' next...
EDIT
Ok, here's the next step, I think, decoupling memoization from factorial:
EDIT
Ok, here's a fully generalized version that seems to work.
当然,记忆尾递归函数的困境是,当尾递归函数
调用自身时,不允许它对递归调用的结果执行任何操作,包括将其放入缓存中。棘手;那么我们能做什么呢?
这里的关键见解是,由于不允许递归函数对递归调用的结果执行任何操作,因此递归调用的所有参数的结果将是相同的!因此,如果递归调用跟踪是这样的
,那么对于 x0,x1,...,xN 中的所有 x,
f x
的结果将是相同的,即 res。因此,递归函数的最后一次调用(即非递归调用)知道所有先前值的结果 - 它可以缓存它们。您唯一需要做的就是将已访问值的列表传递给它。以下是它可能寻找阶乘的内容:但是等等!看 -
fact0
的l
参数包含对fact0
递归调用的所有参数 - 就像非尾递归版本中的堆栈一样!这是完全正确的。任何非尾递归算法都可以通过将“堆栈帧列表”从堆栈移动到堆并将递归调用结果的“后处理”转换为对该数据结构的遍历来转换为尾递归算法。实用说明:上面的阶乘示例说明了一种通用技术。它毫无用处——对于阶乘函数来说,缓存顶级事实 n 的结果就足够了,因为特定 n 的事实 n 的计算只会命中 a fact0 的唯一一系列 (n,res) 参数对 - 如果 (n,1) 尚未缓存,则不会调用任何对fact0。
请注意,在这个例子中,当我们从非尾递归阶乘到尾递归阶乘时,我们利用了乘法是结合性和交换性这一事实——尾递归阶乘执行与非尾递归阶乘不同的乘法集。递归一。
事实上,存在从非尾递归算法到尾递归算法的通用技术,这产生了相当于 Tee 的算法。这种技术称为“连续传递变换”。按照这条路线,您可以采用非尾递归记忆阶乘,并通过几乎机械转换获得尾递归记忆阶乘。请参阅 Brian 的回答以了解此方法的说明。
The predicament of memoizing tail-recursive functions is, of course, that when tail-recursive function
calls itself, it is not allowed to do anything with a result of the recursive call, including putting it into cache. Tricky; so what can we do?
The critical insight here is that since the recursive function is not allowed to do anything with a result of recursive call, the result for all arguments to recursive calls will be the same! Therefore if recursion call trace is this
then for all x in x0,x1,...,xN the result of
f x
will be the same, namely res. So the last invocation of a recursive function, the non-recursive call, knows the results for all the previous values - it is in a position to cache them. The only thing you need to do is to pass a list of visited values to it. Here is what it might look for factorial:But wait! Look -
l
parameter offact0
contains all the arguments to recursive calls tofact0
- just like the stack would in a non-tail-recursive version! That is exactly right. Any non-tail recursive algorithm can be converted to a tail-recursive one by moving the "list of stack frames" from stack to heap and converting the "postprocessing" of recursive call result into a walk over that data structure.Pragmatic note: The factorial example above illustrates a general technique. It is quite useless as is - for factorial function it is quite enough to cache the top-level
fact n
result, because calculation offact n
for a particular n only hits a unique series of (n,res) pairs of arguments to fact0 - if (n,1) is not cached yet, then none of the pairs fact0 is going to be called on are.Note that in this example, when we went from non-tail-recursive factorial to a tail-recursive factorial, we exploited the fact that multiplication is associative and commutative - tail-recursive factorial execute a different set of multiplications than a non-tail-recursive one.
In fact, a general technique exists for going from non-tail-recursive to tail-recursive algorithm, which yields an algorithm equivalent to a tee. This technique is called "continuatuion-passing transformation". Going that route, you can take a non-tail-recursive memoizing factorial and get a tail-recursive memoizing factorial by pretty much a mechanical transformation. See Brian's answer for exposition of this method.
我不确定是否有更简单的方法来做到这一点,但一种方法是创建一个记忆 y 组合器:
然后,您可以使用此组合器代替“let rec”,第一个参数代表函数递归调用:
编辑
正如 Mitya 指出的,
memoY
不保留 memoee 的尾递归属性。这是一个修改后的组合器,它使用异常和可变状态来记忆任何递归函数而不会溢出堆栈(即使原始函数本身不是尾递归!):不幸的是,插入到每个递归调用中的机制有些繁重,因此性能在需要深度递归的未记忆输入上可能会有点慢。然而,与其他一些解决方案相比,这有一个好处,即它需要对递归函数的自然表达进行相当小的更改:
编辑
我将稍微介绍一下它与其他解决方案的比较。该技术利用了异常提供旁路这一事实:
'a -> 类型的函数。 'b
实际上不需要返回'b
类型的值,但可以通过异常退出。如果返回类型明确包含指示失败的附加值,我们就不需要使用异常。当然,我们可以使用'b选项
作为函数的返回类型来达到此目的。这将导致以下记忆组合器:以前,我们的记忆过程如下所示:
现在,我们需要考虑以下事实:
fib
应该返回int option
而不是 <代码>int。给定option
类型的合适工作流程,可以写如下:但是,如果我们愿意更改第一个参数的返回类型(从
int
到 < code>int option 在这种情况下),我们也可以一路走下去,只在返回类型中使用延续,就像 Brian 的解决方案一样。这是他的定义的一个变体:同样,如果我们有一个合适的计算表达式来构建 CPS 函数,我们可以像这样定义我们的递归函数:
这与 Brian 所做的完全相同,但我发现这里的语法更简单跟随。为了实现这一点,我们需要以下两个定义:
I'm not sure if there's a simpler way to do this, but one approach would be to create a memoizing y-combinator:
Then, you can use this combinator in lieu of "let rec", with the first argument representing the function to call recursively:
EDIT
As Mitya pointed out,
memoY
doesn't preserve the tail recursive properties of the memoee. Here's a revised combinator which uses exceptions and mutable state to memoize any recursive function without overflowing the stack (even if the original function is not itself tail recursive!):Unfortunately, the machinery which is inserted into each recursive call is somewhat heavy, so performance on un-memoized inputs requiring deep recursion can be a bit slow. However, compared to some other solutions, this has the benefit that it requires fairly minimal changes to the natural expression of recursive functions:
EDIT
I'll expand a bit on how this compares to other solutions. This technique takes advantage of the fact that exceptions provide a side channel: a function of type
'a -> 'b
doesn't actually need to return a value of type'b
, but can instead exit via an exception. We wouldn't need to use exceptions if the return type explicitly contained an additional value indicating failure. Of course, we could use the'b option
as the return type of the function for this purpose. This would lead to the following memoizing combinator:Previously, our memoization process looked like:
Now, we need to incorporate the fact that
fib
should return anint option
instead of anint
. Given a suitable workflow foroption
types, this could be written as follows:However, if we're willing to change the return type of the first parameter (from
int
toint option
in this case), we may as well go all the way and just use continuations in the return type instead, as in Brian's solution. Here's a variation on his definitions:And again, if we have a suitable computation expression for building CPS functions, we can define our recursive function like this:
This is exactly the same as what Brian has done, but I find the syntax here is easier to follow. To make this work, all we need are the following two definitions:
我写了一个测试来可视化记忆。每个点都是一个递归调用。
kvb 的解决方案返回相同的结果,就像这个函数一样直接记忆。
测试源代码。
I wrote a test to visualize the memoization. Each dot is a recursive call.
kvb's solution returns the same results are straight memoization like this function.
Test source code.
如果通过 y 的相互尾递归没有创建堆栈帧,那么这应该有效:
That should work if mutual tail recursion through y are not creating stack frames: