Lambda lifting: Difference between revisions
mNo edit summary |
(Fixing link) |
||
Line 71: | Line 71: | ||
[[Reducible expression]]. | [[Reducible expression]]. | ||
This converse of lambda lifting is [[ | This converse of lambda lifting is [[lambda dropping]], also known as | ||
avoiding parameter passing | [[avoiding parameter passing]]. | ||
[[Category:Refactoring]] | [[Category:Refactoring]] |
Revision as of 06:18, 20 July 2010
Turning free variables into arguments.
As an example, consider the following Worker wrapper function, which computes the truncated square root of an integer:
isqrt :: Integer -> Integer
isqrt n
| n < 0 = error "isqrt"
| otherwise = isqrt' ((n+1) `div` 2)
where
isqrt' s
| s*s <= n && n < (s+1)*(s+1) = s
| otherwise = isqrt' ((s + (n `div` s)) `div` 2)
Suppose that you find that the worker function, isqrt'
, might be useful somewhere else (say, in a context where a better initial estimate is known). You would like to use the Lifting pattern to raise it to the top level. However, isqrt'
contains a Free variable, n
, which is bound in the outer function. What to do?
The solution is to promote n
to be an argument of isqrt'
. The refactored code might look like this:
isqrt :: Integer -> Integer
isqrt n
| n < 0 = error "isqrt"
| otherwise = isqrt' n ((n+1) `div` 2)
where
isqrt' n s
| s*s <= n && n < (s+1)*(s+1) = s
| otherwise = isqrt' n ((s + (n `div` s)) `div` 2)
The isqrt' function may now be safely lifted to the top-level.
Note that naive lambda lifting can cause a program to be less lazy. Consider, for example:
f x y = g x + g (2*x)
where
g x = sqrt y + x
If you want to lift the definition of g
, you might be tempted to write:
f x y = g y x + g y (2*x)
where
g y x = sqrt y + x
However, this would mean that sqrt y
is evaluated twice, whereas in the first program, it would be evaluated only once. A more efficient approach is not to lift out the variable y
, but rather the expression sqrt y
:
f x y = let sy = sqrt y in g sy x + g sy (2*x)
where
g sy x = sy + x
An expression of this sort which only mentions Free variables is
called a Free expression. If a free expression is as large as it can
be, it is called a Maximal free expression, or MFE for short. Note that
sqrt y
is actually not technically maximal. Lifting out the MFE would give you:
f x y = let psy = (+) (sqrt y) in g psy x + g psy (2*x)
where
g psy x = psy x
However, you save no more work here than the second version, and in addition, the resulting function is harder to read. In general, it only makes sense to abstract out a Free expression if it is also a Reducible expression.
This converse of lambda lifting is lambda dropping, also known as avoiding parameter passing.