Sommaire |
If I have this constant pascal defined as
pascal :: [[Int]] pascal = iterate newrow [1] where newrow = (zipWith (+) <*> tail) . ([0]++) . (++[0])
And I evaluate pascal !! 50 !! 50 in GHCI, how much of the triangle does this evaulate? Does that laziness mean only the necessary values (plus a bunch of thunks) get calculated?
Yes, only the elements that are required to compute the element in question are evaluated.
GHCi offers the :sprint and :print debugging commands that can give you some information about what parts of a value have already been evaluated.
On this example:
GHCi> :sprint pascal pascal = _
(That s because nothing has been evaluated at this point, and thunks are shown as _.)
GHCi> pascal !! 5 !! 5 1
(I m not using 50 because then this example would become too long.)
GHCi> :sprint pascal
pascal = [1] : [_,1] : [_,_,1] : [_,_,_,1] : [_,_,_,_,1] :
(_ : _ : _ : _ : _ : 1 : _) : _
Now you get a pretty clear idea what parts have been looked at.
Let s try one more:
GHCi> pascal !! 5 !! 4
5
GHCi> :sprint pascal
pascal = [1] : [1,1] : [_,2,1] : [_,_,3,1] : [_,_,_,4,1] :
(_ : _ : _ : _ : 5 : 1 : _) : _
And another one:
GHCi> pascal !! 10 !! 5
252
GHCi> :sprint pascal
pascal = [1] : [1,1] : [1,2,1] : [1,3,3,1] : [1,4,6,4,1] :
(1 : 5 : 10 : 10 : 5 : 1 : _) :
(_ : 6 : 15 : 20 : 15 : 6 : _) :
(_ : _ : 21 : 35 : 35 : 21 : _) :
(_ : _ : _ : 56 : 70 : 56 : _) :
(_ : _ : _ : _ : 126 : 126 : _) :
(_ : _ : _ : _ : _ : 252 : _) : _
License : cc by-sa 3.0
http://stackoverflow.com/questions/26729146/how-much-of-pascals-triangle-does-this-evaluate
