Haskell Decimal to Binary - haskell

I am trying to build a function that converts a Decimal(Int) into a Binary number.
Unfortunately other than in java it is not possible to divide an int by two in haskell.
I am very new to functional programming so the problem could be something trivial.
So far I could not find another solution to this problem but
here is my first try :
fromDecimal :: Int -> [Int]
fromDecimal 0 = [0]
fromDecimal n = if (mod n 2 == 0) then
do
0:fromDecimal(n/2)
else
do
1:fromDecimal(n/2)
I got an java implementation here which I did before :
public void fromDecimal(int decimal){
for (int i=0;i<values.length;i++){
if(decimal % 2 = 0)
values[i]=true ;
decimal = decimal/ 2;
else {values[i]= false;
} }
}
Hopefully this is going to help to find a solution!

There are some problems with your solution. First of all, I advise not to use do at all, until you understand what do does. Here we do not need do at all.
Unfortunately other than in java it is not possible to divide an int by two in haskell.
It actually is, but the / operator (which is in fact the (/) function), has type (/) :: Fractional a => a -> a -> a. An Int is not Fractional. You can perform integer division with div :: Integral a => a -> a -> a.
So then the code looks like:
fromDecimal :: Int -> [Int]
fromDecimal 0 = [0]
fromDecimal n = if (mod n 2 == 0) then 0:fromDecimal (div n 2) else 1:fromDecimal (div n 2)
But we can definitely make this more elegant. mod n 2 can only result in two outcomes: 0 and 1, and these are exactly the ones that we use at the left side of the (:) operator.
So we do not need to use an if-then-else at all:
fromDecimal :: Int -> [Int]
fromDecimal 0 = [0]
fromDecimal n = mod n 2 : fromDecimal (div n 2)
Likely this is still not exactly what you want: here we write the binary value such that the last element, is the most significant one. This function will add a tailing zero, which does not make a semantical difference (due to that order), but it is not elegant either.
We can define an function go that omits this zero, if the given value is not zero, like:
fromDecimal :: Int -> [Int]
fromDecimal 0 = [0]
fromDecimal n = go n
where go 0 = []
go k = mod k 2 : go (div k 2)
If we however want to write the most significant bit first (so in the same order as we write decimal numbers), then we have to reverse the outcome. We can do this by making use of an accumulator:
fromDecimal :: Int -> [Int]
fromDecimal 0 = [0]
fromDecimal n = go n []
where go 0 r = r
go k rs = go (div k 2) (mod k 2:rs)

You cannot / integers in Haskell – division is not defined in terms of integral numbers! For integral division use div function, but in your case more suitable would be divMod that comes with mod gratis.
Also, you are going to get reversed output, so you can reverse manually it after that, or use more memory-efficient version with accumulator:
decToBin :: Int -> [Int]
decToBin = go [] where
go acc 0 = acc
go acc n = let (d, m) = n `divMod` 2 in go (m : acc) d
go will give you an empty list for 0. You may add it manually if the list is empty:
decToBin = (\l -> if null l then [0] else l) . go [] where ...

Think through how your algorithm will work. It starts from 2⁰, so it will generate bits backward from how we ordinarily think of them, i.e., least-significant bit first. Your algorithm can represent non-negative binary integers only.
fromDecimal :: Int -> [Int]
fromDecimal d | d < 0 = error "Must be non-negative"
| d == 0 = [0]
| otherwise = reverse (go d)
where go 0 = []
go d = d `rem` 2 : go (d `div` 2)
In Haskell, when we generate a list in reverse, go ahead and do so but then reverse the result at the end. The reason for this is consing up a list (gluing new items at the head with :) has a constant cost and the reverse at the end has a linear cost — but appending with ++ has a quadratic cost.
Common Haskell style is to have a private inner loop named go that the outer function applies when it’s happy with its arguments. The base case is to terminate with the empty list when d reaches zero. Otherwise, we take the current remainder modulo 2 and then proceed with d halved and truncated.
Without the special case for zero, fromDecimal 0 would be the empty list rather than [0].

The binary numbers are usually strings and not really used in calculations.
Strings are also less complicated.
The pattern of binary numbers is like any other. It repeats but at a faster clip.
Only a small set is necessary to generate up to 256 (0-255) binary numbers.
The pattern can systematically be expanded for more.
The starting pattern is 4, 0-3
bd = ["00","01","10","11"]
The function to combine them into larger numbers is
d2b n = head.drop n $ [ d++e++f++g | d <- bd, e <- bd, f <- bd, g <- bd]
d2b 125
"01111101"
If it's not obvious how to expand, then
bd = ["000","001","010","011","100","101","110","111"]
Will give you up to 4096 binary digits (0-4095). All else stays the same.
If it's not obvious, the db2 function uses 4 pairs of binary numbers so 4 of the set. (2^8) - 1 or (2^12) - 1 is how many you get.
By the way, list comprehension are sugar coated do structures.
Generate the above patterns with
[ a++b | a <- ["0","1"], b <- ["0","1"] ]
["00","01","10","11"]
and
[ a++b++c | a <- ["0","1"], b <- ["0","1"], c <- ["0","1"] ]
["000","001","010","011","100","101","110","111"]
More generally, one pattern and one function may serve the purpose
b2 = ["0","1"]
b4 = [ a++b++c++d | a <- b2, b <- b2, c <- b2, d <- b2]
b4
["0000","0001","0010","0011","0100","0101","0110","0111","1000","1001","1010","1011","1100","1101","1110","1111"]
bb n = head.drop n $ [ a++b++c++d | a <- b4, b <- b4, c <- b4, d <- b4]
bb 32768
"1000000000000000"
bb 65535
"1111111111111111"
To calculate binary from decimal directly in Haskell using subtraction
cvtd n (x:xs) | x>n = 0:(cvtd n xs)
| n>x = 1:(cvtd (n-x) xs)
| True = 1:[0|f<-xs]
Use any number of bits you want, for example 10 bits.
cvtd 639 [2^e|e<-[9,8..0]]
[1,0,0,1,1,1,1,1,1,1]

import Data.List
dec2bin x =
reverse $ binstr $ unfoldr ndiv x
where
binstr = map (\x -> "01" !! x)
exch (a,b) = (b,a)
ndiv n =
case n of
0 -> Nothing
_ -> Just $ exch $ divMod n 2

Related

Alternative way to write a Haskell function of single input and output of this kind

Good morning everyone!
I'm using the following function as a fitting example of a function that needs to have a simple input and output. In this case it's a function that converts a number from decimal to binary form, as a list of digits no less, just because it is convenient later on.
I chose to write it like this, because even though a number goes in and a list comes out, another structure is needed as an intermediate step, that will hold the digits found so far and hold the quotient of the division, as input for the next step of the loop. I will clean up the necessary mess before outputing anything, though, by selecting the part of the structure that I'm interested in, in this case the second one , and not counters or other stuff, that I'm done with. (As I mentioned this is an example only, and it's not unusual in other cases to initialize the until loop with a triplet like (a,b,c), only to pick one of them at the end, as I see fit, with the help of additional function, like pickXof3.)
So there,
dec2Bin :: Int -> [Int]
dec2Bin num = snd $ until
(\(n,l) -> n <=0) -- test
(\(n,l) -> (fst $ division n, (snd $ division n):l)) -- main function
(num,[]) -- initialization
where division a = divMod a 2
I find it very convenient that Haskell, although lacking traditional for/while loops has a function like until, which reminds me very much of Mathematica's NextWhile, that I'm familiar with.
In the past I would write sth even uglier, like two functions, a "helper" one and a "main" one, like so
dec2BinHelper :: (Int,[Int]) -> (Int,[Int])
dec2BinHelper (n,l)
| n <= 0 = (n,l)
| otherwise = dec2BinHelper (fst $ division n, (snd $ division n):l)
where division a = divMod a 2
-- a function with the sole purpose to act as a front-end to the helper function, initializing its call parameters and picking up its output
dec2Bin :: Int -> [Int]
dec2Bin n = snd $ dec2BinHelper (n,[])
which I think is unnecessarily bloated.
Still, while the use of until allows me to define just one function, I get the feeling that it could be done even simpler/easier to read, perhaps in a way more fitting to functional programming. Is that so? How would you write such a function differently, while keeping the input and output at the absolutely essential values?
I strongly prefer your second solution. I'd start a clean-up with two things: use pattern matching, and use where to hide your helper functions. So:
dec2Bin :: Int -> [Int]
dec2Bin n = snd $ dec2BinHelper (n, []) where
dec2BinHelper (n, l)
| n <= 0 = (n, l)
| otherwise = dec2BinHelper (d, m:l)
where (d, m) = divMod n 2
Now, in the base case, you return a tuple; but then immediately call snd on it. Why not fuse the two?
dec2Bin :: Int -> [Int]
dec2Bin n = dec2BinHelper (n, []) where
dec2BinHelper (n, l)
| n <= 0 = l
| otherwise = dec2BinHelper (d, m:l)
where (d, m) = divMod n 2
There's no obvious reason why you should pass these arguments in a tuple, rather than as separate arguments, which is more idiomatic and saves some allocation/deallocation noise besides.
dec2Bin :: Int -> [Int]
dec2Bin n = dec2BinHelper n [] where
dec2BinHelper n l
| n <= 0 = l
| otherwise = dec2BinHelper d (m:l)
where (d, m) = divMod n 2
You can swap the arguments to dec2BinHelper and eta-reduce; that way, you will not be shadowing the definition of n.
dec2Bin :: Int -> [Int]
dec2Bin = dec2BinHelper [] where
dec2BinHelper l n
| n <= 0 = l
| otherwise = dec2BinHelper (m:l) d
where (d, m) = divMod n 2
Since you know that n > 0 in the recursive call, you can use the slightly faster quotRem in place of divMod. You could also consider using bitwise operations like (.&. 1) and shiftR 1; they may be even better, but you should benchmark to know for sure.
dec2Bin :: Int -> [Int]
dec2Bin = dec2BinHelper [] where
dec2BinHelper l n
| n <= 0 = l
| otherwise = dec2BinHelper (r:l) q
where (q, r) = quotRem n 2
When you don't have a descriptive name for your helper function, it's traditional to name it go or loop.
dec2Bin :: Int -> [Int]
dec2Bin = go [] where
go l n
| n <= 0 = l
| otherwise = go (r:l) q
where (q, r) = quotRem n 2
At this point, the two sides of the conditional are short enough that I'd be tempted to put them on their own line, though this is something of an aesthetic choice.
dec2Bin :: Int -> [Int]
dec2Bin = go [] where
go l n = if n <= 0 then l else go (r:l) q
where (q, r) = quotRem n 2
Finally, a comment on the name: the input isn't really in decimal in any meaningful sense. (Indeed, it's much more physically accurate to think of the input as already being in binary!) Perhaps int2Bin or something like that would be more accurate. Or let the type speak for itself, and just call it toBin.
toBin :: Int -> [Int]
toBin = go [] where
go l n = if n <= 0 then l else go (r:l) q
where (q, r) = quotRem n 2
At this point I'd consider this code quite idiomatic.

Haskell compare issue

I use recursion to compare two numbers(ex 123 is the same with 123) and store how many digits they have in common(ex 123 compared to 123 has 3 and with 124 has 2).
While my program does find that number i want to put conditions for every situation (ex if they have 2 digits in common output value 44 or if they have 3 do something else etc.) but no comparison is done. Can someone please explain to me what and why that happens. Here my code :
dg :: Int->Int->Int
dg 0 0 = 0
dg x y = if (c==2) then 23 else 24 -- c = common digits
where c = digits (x `div` 10) (y `div` 10) + if (x `mod` 10 == y `mod` 10) then 1 else 0
I run hugs deleting the "if" condition, giving input 10 10 the output is 2 but when i leave it on and rerun it, it goes to 24 (while 23 is correct). I'm really confused.
To find how many digits two Int have in common, using the string representation isn't too bad (It's actually terrible, see update below).
commonDigits :: Int -> Int -> Int
commonDigits a b = length . filter id $ zipWith (==) (reverse $ show a) (reverse $ show b)
reverse is needed to make sure the digits align properly.
If you're into point-free style:
import Data.Function
commonDigits :: Int -> Int -> Int
commonDigits = fmap (length . filter id) . zipWith (==) `on` reverse . show
Then if you want to return a special Int depending on the result of commonDigits you can use a separate function:
specialResult :: Int -> Int -> Int
specialResult a b =
case commonDigits a b of
2 -> 23
3 -> 48
_ -> 256
Update: For negative integers this method is not good. If only one of the argument is negative this function should behave like your mod 10 version. If both are negative there is the possibility that the '-' sign is counted as a digit.
Let's use a digits function that will give a list of digits for both positive and negative numbers from the least significant to the most significant digits. What happens if the number 0 is given as input, should we return an empty list or [0]? Your call on that one, but I'm going to assume you want [0].
digits :: Int -> [Int]
digits 0 = [0]
digits n = digits' (abs n)
where
digits' 0 = []
digits' n = n `mod` 10 : digits' (n `div` 10)
With this new digits function we can rewrite commonDigits to:
commonDigits :: Int -> Int -> Int
commonDigits = fmap (length . filter id) . zipWith (==) `on` digits

Solve the equation a * b = c, where a, b and c are natural numbers

I have some natural number c. I want to find all pairs of natural numbers a and b, where a < b, such as a * b = c.
I have a solution:
solve c = do solveHelper [1..c] c where
solveHelper xs c = do
x <- xs
(division, modulo ) <- return (c `divMod` x)
True <- return (modulo == 0)
True <- return (x <= division)
return (x, division)
Example:
*Main> solve 10
[(1,10),(2,5)]
Is there a way to accelerate my code, or a better algorithm I should use?
You can do much, much better. The basic idea is this: first, factorize the number; then enumerate the partitions of the factorization. The product of each partition is a solution. There are fast factorization algorithms out there, but even the naive one is quite an improvement on your code; so:
factorize :: Integer -> [Integer]
factorize n
| n < 1 = error "no. =("
| otherwise = go 2 n
where
go p n | p * p > n = [n]
go p n = case quotRem n p of
(q, 0) -> p:go p q
_ -> go (p+1) n
I will use the very nice multiset-comb package to compute partitions of the set of factors. It doesn't support the usual Foldable/Traversable stuff out of the box, so we have to roll our own product operation -- but in fact this can be a bit more efficient than using the product that the standard interface would give us anyway.
import Math.Combinatorics.Multiset
productMS :: Multiset Integer -> Integer
productMS (MS cs) = product [n^p | (n, p) <- cs]
divisors :: Integer -> [(Integer, Integer)]
divisors n =
[ (a, b)
| (aMS, bMS) <- splits (fromList (factorize n))
, let a = productMS aMS; b = productMS bMS
, a <= b
]
For unfair timings, we can compare in ghci:
*Main> :set +s
*Main> length $ solve (product [1..10])
135
(3.55 secs, 2,884,836,952 bytes)
*Main> length $ divisors (product [1..10])
135
(0.00 secs, 4,612,104 bytes)
*Main> length $ solve (product [1..15])
^CInterrupted. [after several minutes, I gave up]
*Main> length $ divisors (product [1..15])
2016
(0.03 secs, 33,823,168 bytes)
Here solve is your solution, divisors is mine. For a fair comparison, we should compile; I used this program:
main = print . last . solve . product $ [1..11]
(And similar with divisors in place of solve.) I compiled with -O2; yours used 1.367s total, mine 0.002s total.
There's one optimization you don't use: you don't have to try every value from 0 to c.
a < b and a * b = c, so a * a < c, meaning you only have to try numbers from 0 to sqrt c. Or, if you don't want to compute the square root of c, you can stop as soon as a * a >= c.
To do so, you can replace [1..c] by (takeWhile (\x -> x * x < c) [1..]).

Haskell reverse Integer with recursion

I want to reverse an Integer in Haskell with recursion. I have a small issue.
Here is the code :
reverseInt :: Integer -> Integer
reverseInt n
| n>0 = (mod n 10)*10 + reverseInt(div n 10)
| otherwise = 0
Example 345
I use as input 345 and I want to output 543
In my program it will do....
reverseInt 345
345>0
mod 345 10 -> 5
reverseInt 34
34
34>0
mod 34 10 -> 4
reverseInt 3
3>0
mod 3 10 -> 3
reverseInt 0
0=0 (ends)
And at the end it returns the sum of them... 5+4+3 = 12.
So I want each time before it sums them, to multiple the sum * 10. So it will go...
5
5*10 + 4
54*10 + 3
543
Here's a relatively simple one:
reverseInt :: Int -> Int
reverseInt 0 = 0
reverseInt n = firstDigit + 10 * (reverseInt $ n - firstDigit * 10^place)
where
n' = fromIntegral n
place = (floor . logBase 10) n'
firstDigit = n `div` 10^place
Basically,
You take the logBase 10 of your input integer, to give you in what place it is (10s, 100s, 1000s...)
Because the previous calculation gives you a floating point number, of which we do not need the decimals, we use the floor function to truncate everything after the decimal.
We determine the first digit of the number by doing n 'div' 10^place. For example, if we had 543, we'd find place to be 2, so firstDigit = 543/100 = 5 (integer division)
We use this value, and add it to 10 * the reverse of the 'rest' of the integer, in this case, 43.
Edit: Perhaps an even more concise and understandable version might be:
reverseInt :: Int -> Int
reverseInt 0 = 0
reverseInt n = mod n 10 * 10^place + reverseInt (div n 10)
where
n' = fromIntegral n
place = (floor . logBase 10) n'
This time, instead of recursing through the first digit, we're recursing through the last one and using place to give it the right number of zeroes.
reverseInt :: Integer -> Integer
reverseInt n = snd $ rev n
where
rev x
| x>0 = let (a,b) = rev(div x 10)
in ((a*10), (mod x 10)*a + b)
| otherwise = (1,0)
Explanation left to reader :)
I don't know convenient way to found how many times you should multiply (mod n 10) on 10 in your 3rd line. I like solution with unfoldr more:
import Data.List
listify = unfoldr (\ x -> case x of
_ | x <= 0 -> Nothing
_ -> Just(mod x 10, div x 10) )
reverse_n n = foldl (\ acc x -> acc*10+x) 0 (listify n)
In listify function we generate list of numbers from integer in reverse order and after that we build result simple folding a list.
Or just convert it to a string, reverse it and convert it back to an integer:
reverseInt :: Integer -> Integer
reverseInt = read . reverse . show
More (not necessarily recursion based) answers for great good!
reverseInt 0 = 0
reverseInt x = foldl (\x y -> 10*x + y) 0 $ numToList x
where
numToList x = if x == 0 then [] else (x `rem` 10) : numToList (x `div` 10)
This is basically the concatenation of two functions : numToList (convert a given integer to a list 123 -> [1,2,3]) and listToNum (do the opposite).
The numToList function works by repeatedly getting the lowest unit of the number (using rem, Haskell's remainder function), and then chops it off (using div, Haskell's integer division function). Once the number is 0, the empty list is returned and the result concatenates into the final list. Keep in mind that this list is in reverse order!
The listToNum function (not seen) is quite a sexy piece of code:
foldl (\x y -> 10*x + y) 0 xs
This starts from the left and moves to the right, multiplying the current value at each step by 10 and then adding the next number to it.
I know the answer has already been given, but it's always nice to see alternative solutions :)
The first function is recursive to convert the integer to a list. It was originally reversing but the re-conversion function reversed easier so I took it out of the first. The functions can be run separately. The first outputs a tuple pair. The second takes a tuple pair. The second is not recursive nor did it need to be.
di 0 ls = (ls,sum ls); di n ls = di nn $ d:ls where (nn,d) = divMod n 10
di 3456789 []
([3,4,5,6,7,8,9],42)
rec (ls,n) = (sum [y*(10^x)|(x,y) <- zip [0..] ls ],n)
Run both as
rec $ di 3456789 []
(9876543,42)

Euler #4 with bigger domain

Consider the modified Euler problem #4 -- "Find the maximum palindromic number which is a product of two numbers between 100 and 9999."
rev :: Int -> Int
rev x = rev' x 0
rev' :: Int -> Int -> Int
rev' n r
| n == 0 = r
| otherwise = rev' (n `div` 10) (r * 10 + n `mod` 10)
pali :: Int -> Bool
pali x = x == rev x
main :: IO ()
main = print . maximum $ [ x*y | x <- nums, y <- nums, pali (x*y)]
where
nums = [9999,9998..100]
This Haskell solution using -O2 and ghc 7.4.1 takes about 18
seconds.
The similar C solution takes 0.1 second.
So Haskell is 180 times
slower. What's wrong with my solution? I assume that this type of
problems Haskell solves pretty well.
Appendix - analogue C solution:
#define A 100
#define B 9999
int ispali(int n)
{
int n0=n, k=0;
while (n>0) {
k = 10*k + n%10;
n /= 10;
}
return n0 == k;
}
int main(void)
{
int max = 0;
for (int i=B; i>=A; i--)
for (int j=B; j>=A; j--) {
if (i*j > max && ispali(i*j))
max = i*j; }
printf("%d\n", max);
}
The similar C solution
That is a common misconception.
Lists are not loops!
And using lists to emulate loops has performance implications unless the compiler is able to eliminate the list from the code.
If you want to compare apples to apples, write the Haskell structure more or less equivalent to a loop, a tail recursive worker (with strict accumulator, though often the compiler is smart enough to figure out the strictness by itself).
Now let's take a more detailed look. For comparison, the C, compiled with gcc -O3, takes ~0.08 seconds here, the original Haskell, compiled with ghc -O2 takes ~20.3 seconds, with ghc -O2 -fllvm ~19.9 seconds. Pretty terrible.
One mistake in the original code is to use div and mod. The C code uses the equivalent of quot and rem, which map to the machine division instructions and are faster than div and mod. For positive arguments, the semantics are the same, so whenever you know that the arguments are always non-negative, never use div and mod.
Changing that, the running time becomes ~15.4 seconds when compiling with the native code generator, and ~2.9 seconds when compiling with the LLVM backend.
The difference is due to the fact that even the machine division operations are quite slow, and LLVM replaces the division/remainder with a multiply-and-shift operation. Doing the same by hand for the native backend (actually, a slightly better replacement taking advantage of the fact that I know the arguments will always be non-negative) brings its time down to ~2.2 seconds.
We're getting closer, but are still a far cry from the C.
That is due to the lists. The code still builds a list of palindromes (and traverses a list of Ints for the two factors).
Since lists cannot contain unboxed elements, that means there is a lot of boxing and unboxing going on in the code, that takes time.
So let us eliminate the lists, and take a look at the result of translating the C to Haskell:
module Main (main) where
a :: Int
a = 100
b :: Int
b = 9999
ispali :: Int -> Bool
ispali n = go n 0
where
go 0 acc = acc == n
go m acc = go (m `quot` 10) (acc * 10 + (m `rem` 10))
maxpal :: Int
maxpal = go 0 b
where
go mx i
| i < a = mx
| otherwise = go (inner mx b) (i-1)
where
inner m j
| j < a = m
| p > m && ispali p = inner p (j-1)
| otherwise = inner m (j-1)
where
p = i*j
main :: IO ()
main = print maxpal
The nested loop is translated to two nested worker functions, we use an accumulator to store the largest palindrome found so far. Compiled with ghc -O2, that runs in ~0.18 seconds, with ghc -O2 -fllvm it runs in ~0.14 seconds (yes, LLVM is better at optimising loops than the native code generator).
Still not quite there, but a factor of about 2 isn't too bad.
Maybe some find the following where the loop is abstracted out more readable, the generated core is for all intents and purposes identical (modulo a switch of argument order), and the performance of course the same:
module Main (main) where
a :: Int
a = 100
b :: Int
b = 9999
ispali :: Int -> Bool
ispali n = go n 0
where
go 0 acc = acc == n
go m acc = go (m `quot` 10) (acc * 10 + (m `rem` 10))
downto :: Int -> Int -> a -> (a -> Int -> a) -> a
downto high low acc fun = go high acc
where
go i acc
| i < low = acc
| otherwise = go (i-1) (fun acc i)
maxpal :: Int
maxpal = downto b a 0 $ \m i ->
downto b a m $ \mx j ->
let p = i*j
in if mx < p && ispali p then p else mx
main :: IO ()
main = print maxpal
#axblount is at least partly right; the following modification makes the program run almost three times as fast as the original:
maxPalindrome = foldl f 0
where f a x | x > a && pali x = x
| otherwise = a
main :: IO ()
main = print . maxPalindrome $ [x * y | x <- nums, y <- nums]
where nums = [9999,9998..100]
That still leaves a factor 60 slowdown, though.
This is more true to what the C code is doing:
maxpali :: [Int] -> Int
maxpali xs = go xs 0
where
go [] m = m
go (x:xs) m = if x > m && pali(x) then go xs x else go xs m
main :: IO()
main = print . maxpali $ [ x*y | x <- nums, y <- nums ]
where nums = [9999,9998..100]
On my box this takes 2 seconds vs .5 for the C version.
Haskell may be storing that entire list [ x*y | x <- nums, y <- nums, pali (x*y)] where as the C solution calculates the maximum on the fly. I'm not sure about this.
Also the C solution will only calculate ispali if the product beats the previous maximum. I would bet Haskell calculates are palindrome products regardless of whether x*y is a possible max.
It seems to me that you are having a branch prediction problem. In the C code, you have two nested loops and as soon as a palindrome is seen in the inner loop, the rest of the inner loop will be skipped very fast.
The way you feed this list of products instead of the nested loops I am not sure that ghc is doing any of this prediction.
Another way to write this is to use two folds, instead of one fold over the flattened list:
-- foldl g0 0 [x*y | x<-[b-1,b-2..a], y<-[b-1,b-2..a], pali(x*y)] (A)
-- foldl g1 0 [x*y | x<-[b-1,b-2..a], y<-[b-1,b-2..a]] (B)
-- foldl g2 0 [ [x*y | y<-[b-1,b-2..a]] | x<-[b-1,b-2..a]] (C)
maxpal b a = foldl f1 0 [b-1,b-2..a] -- (D)
where
f1 m x = foldl f2 m [b-1,b-2..a]
where
f2 m y | p>m && pali p = p
| otherwise = m
where p = x*y
main = print $ maxpal 10000 100
Seems to run much faster than (B) (as in larsmans's answer), too (only 3x - 4x slower then the following loops-based code). Fusing foldl and enumFromThenTo definitions gets us the "functional loops" code (as in DanielFischer's answer),
maxpal_loops b a = f (b-1) 0 -- (E)
where
f x m | x < a = m
| otherwise = g (b-1) m
where
g y m | y < a = f (x-1) m
| p>m && pali p = g (y-1) p
| otherwise = g (y-1) m
where p = x*y
The (C) variant is very suggestive of further algorithmic improvements (that's outside the scope of the original Q of course) that exploit the hidden order in the lists, destroyed by the flattening:
{- foldl g2 0 [ [x*y | y<-[b-1,b-2..a]] | x<-[b-1,b-2..a]] (C)
foldl g2 0 [ [x*y | y<-[x, x-1..a]] | x<-[b-1,b-2..a]] (C1)
foldl g0 0 [ safehead 0 . filter pali $
[x*y | y<-[x, x-1..a]] | x<-[b-1,b-2..a]] (C2)
fst $ until ... (\(m,s)-> (max m .
safehead 0 . filter pali . takeWhile (> m) $
head s, tail s))
(0,[ [x*y | y<-[x, x-1..a]] | x<-[b-1,b-2..a]]) (C3)
safehead 0 $ filter pali $ mergeAllDescending
[ [x*y | y<-[x, x-1..a]] | x<-[b-1,b-2..a]] (C4)
-}
(C3) can stop as soon as the head x*y in a sub-list is smaller than the currently found maximum. It is what short-cutting functional loops code could achieve, but not (C4), which is guaranteed to find the maximal palindromic number first. Plus, for list-based code its algorithmic nature is more visually apparent, IMO.

Resources