I'm trying to change a list of Ints in Haskell to make keep them within specific limits, but it doesn't seem to be working. I'm trying to make every int in the list sit between 32 and 127 but it's not working, could anyone explain why this isn't working?
limit :: Int -> Int
limit n | n > 127 = n `mod` 127 + 32
| n < 32 = n + 127 - 32
| otherwise = n
limitList :: [Int] -> [Int]
limitList [] = []
limitList (x:xs) = [limit x] ++ limitList xs
Based on your comment, you want to transform Ints that are not in the 32-127 range by applying a modulo transformation. Therefore we can first implement a helper function:
helper x = 32 + mod (x-32) (128-32)
This results in:
Prelude> helper 31
127
Prelude> helper 32
32
Prelude> helper 127
127
Prelude> helper 128
32
Next our function limitList is only a mapping with that helper:
limitList = map helper
where helper x = 32 + mod (x-32) (128-32)
This generates:
Prelude> limitList [1,4..256]
[97,100,103,106,109,112,115,118,121,124,127,34,37,40,43,46,49,52,55,58,61,64,67,70,73,76,79,82,85,88,91,94,97,100,103,106,109,112,115,118,121,124,127,34,37,40,43,46,49,52,55,58,61,64,67,70,73,76,79,82,85,88,91,94,97,100,103,106,109,112,115,118,121,124,127,34,37,40,43,46,49,52,55,58,61,64]
Related
So I have to make a decimal number into binary list like so: intToBitString 4 = [1,0,0].
Which i have done like so:
intToBitString n = reverse (helper n)
helper 0 = []
helper n
| n `mod` 2 == 1 = 1 : helper (n `div` 2)
| n `mod` 2 == 0 = 0 : helper(n `div` 2)
But then I also have to make a function called intToByte, which pads out the list with 0-s until it's length is 8 elements long. (so making it a bytestring) Like this:
intToByte 7 = [0, 0, 0, 0, 0, 1, 1, 1]
I have tried so many things, but they never work. I am a beginner, so I only know the "if" loop the way I showed above, and recursion, but I dont know anything fancy. One of my tries:
intToByte 0 = [0]
intToByte n
| eight n == helper2 n = reverse (helper2 n)
| otherwise = eight n
helper2 0 = []
helper2 n
| n `mod` 2 == 1 = 1 : helper2 (n `div` 2)
| n `mod` 2 == 0 = 0 : helper2 (n `div` 2)
eight n
| length (helper2 n) < 8 = 0 : eight n
| otherwise = helper2 n
I have been working on this for so many hours that i'm getting confused by it. But this is part of an important assignment, so help would be very appreciated!
First of all, you can simplify your code with:
helper2 :: Integral i => i -> [i]
helper2 0 = []
helper2 n = r : helper2 q
where (q,r) = quotRem n 2
Secondly, the above is a big endian representation [wiki]. Indeed, 7 is represented as [1,1,1], whereas 14 is for example represented as [0,1,1,1]. If we want to revers this, we can work with an accumulator:
helper2 :: Integral i => i -> [i]
helper2 = go []
where go rs 0 = rs
go rs n = go (r:rs) q
where (q,r) = quotRem n 2
This thus maps 7 to [1,1,1] and 14 to [1,1,1,0]. But now we still need to add leading zeros. We can do that for example by maintaing the number of elements already added to the list:
eight :: Integral i => i -> [i]
eight = go [] 0
where go rs l 0 = replicate (8-l) 0 ++ rs
go rs l n = go (r:rs) (l+1) q
where (q,r) = quotRem n 2
Padding can be as simple as computing how many additional elements to push to the list and then have those elements produced using the function replicate from the Prelude:
padLeft :: Int -> a -> [a] -> [a]
padLeft n x xs = replicate (n - length xs) x ++ xs
For instance:
> padLeft 8 0 [1, 1, 0]
[0,0,0,0,0,1,1,0]
One approach would be to define a function bits such that bits k converts its argument to a bit string of length k:
bits :: Int -> Int -> [Int]
bits 0 _n = []
bits k n | n < 0 = error "bits: negative"
| n > 2 * m - 1 = error "bits: overflow"
| otherwise = let (i, j) = n `divMod` m in i : bits (k - 1) j
where m = 2 ^ (k - 1)
Your function eight is then easily written as
eight :: Int -> [Int]
eight = bits 8
This gives:
> eight 4
[0,0,0,0,0,1,0,0]
> eight 7
[0,0,0,0,0,1,1,1]
I´m new to Haskell.
Let´s say I want to sum up the first n elements of a list with a generated function on my own. I don´t know how to do this with Haskell. I just know how to sum up a whole given list, e.g.
sumList :: [Int] -> Int
sumList [] = 0
sumList (x:xs) = x + sumList xs
In order to sum up the first n elements of a list, for example
take the first 5 numbers from [1..10], which is 1+2+3+4+5 = 15
I thought I could do something like this:
sumList :: Int -> [Int] -> Int
sumList take [] = 0
sumList take (x:xs) = x + take $ sumList xs
But it doesn´t work... What´s wrong?
So you know how to sum up the numbers in a list,
sumList :: [Int] -> Int
sumList [] = 0
sumList (x:xs) = x + sumList xs
and if that list has no more than 5 elements in it, this function will even return the correct result if you indeed intended to sum no more than 5 elements in an argument list. Let's make our expectations explicit by renaming this function,
sumUpToFiveElements :: [Int] -> Int
sumUpToFiveElements [] = 0
sumUpToFiveElements (x:xs) = x + sumUpToFiveElements xs
it won't return the correct result for lists longer than five, but at least the name is right.
Can we fix that? Can we count up to 5? Can we count up to 5 while also advancing along the input list as we do?
sumUpToFiveElements :: Int -> [Int] -> Int
sumUpToFiveElements counter [] = 0
sumUpToFiveElements counter (x:xs) = x + sumUpToFiveElements (counter + 1) xs
This still isn't right of course. We do now count, but for some reason we ignore the counter. What is the right time to react to the counter, if we want no more than 5 elements? Let's try counter == 5:
sumUpToFiveElements :: Int -> [Int] -> Int
sumUpToFiveElements 5 [] = 0
sumUpToFiveElements counter [] = 0
sumUpToFiveElements counter (x:xs) = x + sumUpToFiveElements (counter + 1) xs
But why do we demand the list to also be empty when 5 is reached? Let's not do that:
sumUpToFiveElements :: Int -> [Int] -> Int
sumUpToFiveElements 5 _ = 0 -- the wildcard `_` matches *anything*
sumUpToFiveElements counter [] = 0
sumUpToFiveElements counter (x:xs) = x + sumUpToFiveElements (counter + 1) xs
Success! We now stop counting when 5 is reached! More, we also stop the summation!!
Wait, but what was the initial value of counter? We didn't specify it, so it's easy for a user of our function (that would be ourselves) to err and use an incorrect initial value. And by the way, what is the correct initial value?
Okay, so let's do this:
sumUpToFiveElements :: [Int] -> Int
sumUpToFiveElements xs = go 1 xs -- is 1 the correct value here?
where
go counter _ | counter == 5 = 0
go counter [] = 0
go counter (x:xs) = x + go (counter + 1) xs
Now we don't have that extraneous argument that made our definition so brittle, so prone to a user error.
And now for the punchline:
Generalize! (by replacing an example value with a symbolic one; changing 5 to n).
sumUpToNElements :: Int -> [Int] -> Int
sumUpToNElements n xs = .......
........
Done.
One more word of advice: don't use $ while at the very beginning of your learning Haskell. Use explicit parens.
sumList take (x:xs) = x + take $ sumList xs
is parsed as
sumList take (x:xs) = (x + take) (sumList xs)
This adds together two unrelated numbers, and then uses the result as a function to be called with (sumList xs) as an argument (in other words it's an error).
You probably wouldn't write it that way if you were using explicit parens.
Well you should limit the number of values with a parameter (preferably not take, since
that is a function from the Prelude), and thus limit the numbers.
This limiting in your code is apparently take $ sumList xs which is very strange: in your function take is an Int, and $ will basically write your statement to (x + take) (sumList xs). You thus apparently want to perform a function application with (x + take) (an Int) as function, and sumList xs as argument. But an Int is not a function, so it does not typecheck, nor does it include any logic to limit the numbers.
So basically we should consider three cases:
the empty list in which case the sum is 0;
the number of elements to take is less than or equal to zero, in that case the sum is 0; and
the number of elements to take is greater than 0, in that case we add the head to the sum of taking one element less from the tail.
So a straightforward mapping is:
sumTakeList :: (Integral i, Num n) => i -> [n] -> n
sumTakeList _ [] = 0
sumTakeList t (x:xs) | t <= 0 = 0
| otherwise = x + sumTakeList (t-1) xs
But you do not need to write such logic yourself, you can combine the take :: Int -> [a] -> [a] builtin with the sum :: Num a => [a] -> a functions:
sumTakeList :: Num n => Int -> [n] -> n
sumTakeList t = sum . take t
Now if you need to sum the first five elements, we can make that a special case:
subList5 :: Num n => [n] -> n
sumList5 = sumTakeList 5
A great resource to see what functions are available and how they work is Hoogle. Here is its page on take and the documentation for the function you want.
As you can see, the name take is taken, but it is a function you can use to implement this.
Note that your sumList needs another argument, the number of elements to sum. the syntax you want is something like:
sumList :: Int -> [Int] -> Int
sumList n xs = _ $ take n xs
Where the _ are blanks you can fill in yourself. It's a function in the Prelude, but the type signature is a little too complicated to get into right now.
Or you could write it recursively, with two base cases and a third accumulating parameter (by means of a helper function):
sumList :: Int -> [Int] -> Int
sumList n xs = sumList' n xs 0 where
sumList' :: Int -> [Int] -> Int -> Int
sumList' 0 _ a = _ -- A base case.
sumList' _ [] a = _ -- The other base case.
sumList' m (y:ys) a = sumList' _ _ _ -- The recursive case.
Here, the _ symbols on the left of the equals signs should stay there, and mean that the pattern guard ignores that parameter, but the _ symbols on the right are blanks for you to fill in yourself. Again, GHC will tell you the type you need to fill the holes with.
This kind of tail-recursive function is a very common pattern in Haskell; you want to make sure that each recursive call brings you one step closer to the base case. Often, that will mean calling itself with 1 subtracted from a count parameter, or calling itself with the tail of the list parameter as the new list parameter. here, you want to do both. Don't forget to update your running sum, a, when you have the function call itself recursively.
Here's a short-but-sweet answer. You're really close. Consider the following:
The take parameter tells you how many elements you need to sum up, so if you do sumList 0 anything you should always get 0 since you take no elements.
If you want the first n elements, you add the first element to your total and compute the sum of the next n-1 elements.
sumList 0 anything = 0
sumList n [] = 0
sumList n (e:es) = e + sumList (n-1) e
I have following problem:
You are given matrix m*n and you have to find maximal positive ( all elements of submatrix should be > 0) submatrices from (1,1) to (x,y).
What do I mean by maximal is, when you have following matrix:
[[1,2,3,4],[5,6,7,8],[9,10,-11,12],[13,14,15,16]]
then maximal positive submatrices are:
[[[1,2,3,4],[5,6,7,8]],[[1,2],[5,6],[9,10],[13,14]]]
i.e. first two rows is one solution and first two columns is second solution.
Another example: matrix is
[[1,2,3,-4],[5,6,7,8],[-9,10,-11,12],[13,14,15,16]]
and solution is:
[[[1,2,3],[5,6,7]]]
This is my Haskell program which solves it:
import Data.List hiding (insert)
import qualified Data.Set as Set
unique :: Ord a => [a] -> [a]
unique = Set.toList . Set.fromList
subList::[[Int]] ->[[[Int]]]
subList matrix = filter (allPositiveMatrix) $ [ (submatrix matrix 1 1 x y) | x<-[1..width(matrix)], y<-[1..height(matrix)]]
maxWidthMat::[[[Int]]] -> Int
maxWidthMat subList =length ((foldl (\largestPreviousX nextMatrix -> if (length (nextMatrix!!0)) >(length (largestPreviousX !!0)) then nextMatrix else largestPreviousX ) [[]] subList)!!0)
maxWidthSubmatrices:: [[[Int]]] -> Int ->[[[Int]]]
maxWidthSubmatrices subList maxWidth = filter (\x -> (length $x!!0)==maxWidth) subList
height matrix = length matrix
width matrix = length (matrix!!0)
maximalPositiveSubmatrices matrix = maxWidthSubmatrices (subList matrix) (maxWidthMat (filter (\x -> (length $x!!0)==( maxWidthMat $ subList matrix )) (subList matrix)))
allPositiveList list = foldl (\x y -> if (y>0)&&(x==True) then True else False) True list
allPositiveMatrix:: [[Int]] -> Bool
allPositiveMatrix matrix = foldl (\ x y -> if (allPositiveList y)&&(x==True) then True else False ) True matrix
submatrix matrix x1 y1 x2 y2 = slice ( map (\x -> slice x x1 x2) matrix) y1 y2
slice list x y = drop (x-1) (take y list)
maximalWidthSubmatrix mm = maximum $ maximalPositiveSubmatrices mm
maximalHeigthSubmatrix mm = transpose $ maximum $ maximalPositiveSubmatrices $ transpose mm
-- solution
solution matrix =unique $ [maximalWidthSubmatrix matrix]++[maximalHeigthSubmatrix matrix]
As you can see it's extremely lengthy and ugly.
It problably isn't fastest too.
Could you show me more elegant, faster and shorter solution ( possibly with explantions) ?
Proposed algorithm
I think that in order to solve the problem, we first better perform a dimension reduction:
reduce_dim :: (Num a,Ord a) => [[a]] -> [Int]
reduce_dim = map (length . takeWhile (>0)) -- O(m*n)
Here for every row, we calculate the number of items - starting from the left - that are positive. So for the given matrix:
1 2 3 4 | 4
5 6 7 8 | 4
9 10 -11 12 | 2
13 14 15 16 | 4
The second row thus maps to 2, since the third element is -11.
Or for your other matrix:
1 2 3 -4 | 3
5 6 7 8 | 4
-9 10 -11 12 | 0
13 14 15 16 | 4
Since the first row has a -4 at column 4, and the third one at column 1.
Now we can obtain a scanl1 min over these rows:
Prelude> scanl1 min [4,4,2,4] -- O(m)
[4,4,2,2]
Prelude> scanl1 min [3,4,0,4] -- O(m)
[3,3,0,0]
Now each time the number decreases (and at the end), we know we have found a maximal submatrix at the row above. Since that means we now work with a row from where on, the number of columns is less. Once we reach zero, we know that further evaluation has no sense, since we are working with a matrix with 0 columns.
So based on that list, we can simply generate a list of tuples of the sizes of the maximal submatrices:
max_sub_dim :: [Int] -> [(Int,Int)]
max_sub_dim = msd 1 -- O(m)
where msd r [] = []
msd r (0:_) = []
msd r [c] = [(r,c)]
msd r (c1:cs#(c2:_)) | c2 < c1 = (r,c1) : msd (r+1) cs
| otherwise = msd (r+1) cs
So for your two matrices, we obtain:
*Main> max_sub_dim $ scanl1 min $ reduce_dim [[1,2,3,4],[5,6,7,8],[9,10,-11,12],[13,14,15,16]]
[(2,4),(4,2)]
*Main> max_sub_dim $ scanl1 min $ reduce_dim [[1,2,3,-4],[5,6,7,8],[-9,10,-11,12],[13,14,15,16]]
[(2,3)]
Now we only need to obtain these submatrices themselves. We can do this by using take and a map over take:
construct_sub :: [[a]] -> [(Int,Int)] -> [[[a]]]
construct_sub mat = map (\(r,c) -> take r (map (take c) mat)) -- O(m^2*n)
And now we only need to link it all together in a solve:
-- complete program
reduce_dim :: (Num a,Ord a) => [[a]] -> [Int]
reduce_dim = map (length . takeWhile (>0))
max_sub_dim :: [Int] -> [(Int,Int)]
max_sub_dim = msd 1
where msd r [] = []
msd r (0:_) = []
msd r [c] = [(r,c)]
msd r (c1:cs#(c2:_)) | c2 < c1 = (r,c1) : msd (r+1) cs
| otherwise = msd (r+1) cs
construct_sub :: [[a]] -> [(Int,Int)] -> [[[a]]]
construct_sub mat = map (\(r,c) -> take r (map (take c) mat))
solve :: (Num a,Ord a) => [[a]] -> [[[a]]]
solve mat = construct_sub mat $ max_sub_dim $ scanl1 min $ reduce_dim mat
Which then generates:
*Main> solve [[1,2,3,4],[5,6,7,8],[9,10,-11,12],[13,14,15,16]]
[[[1,2,3,4],[5,6,7,8]],[[1,2],[5,6],[9,10],[13,14]]]
*Main> solve [[1,2,3,-4],[5,6,7,8],[-9,10,-11,12],[13,14,15,16]]
[[[1,2,3],[5,6,7]]]
Time complexity
The algorithm runs in O(m×n) with m the number of rows and n the number of columns, to construct the dimensions of the matrices. For every defined function, I wrote the time complexity in comment.
It will take O(m2×n) to construct all submatrices. So the algorithm runs in O(m2×n).
We can transpose the approach and run on columns instead of rows. So in case we are working with matrices where the number of rows differs greatly from the number of columns, we can first calculate the minimum, optionally transpose, and thus make m the smallest of the two.
Point of potential optimization
we can make the algorithm faster by constructing submatrices while constructing max_sub_dim saving some work.
I have a simple function (used for some problems of project Euler, in fact). It turns a list of digits into a decimal number.
fromDigits :: [Int] -> Integer
fromDigits [x] = toInteger x
fromDigits (x:xs) = (toInteger x) * 10 ^ length xs + fromDigits xs
I realized that the type [Int] is not ideal. fromDigits should be able to take other inputs like e.g. sequences, maybe even foldables ...
My first idea was to replace the above code with sort of a "fold with state". What is the correct (= minimal) Haskell-category for the above function?
First, folding is already about carrying some state around. Foldable is precisely what you're looking for, there is no need for State or other monads.
Second, it'd be more natural to have the base case defined on empty lists and then the case for non-empty lists. The way it is now, the function is undefined on empty lists (while it'd be perfectly valid). And notice that [x] is just a shorthand for x : [].
In the current form the function would be almost expressible using foldr. However within foldl the list or its parts aren't available, so you can't compute length xs. (Computing length xs at every step also makes the whole function unnecessarily O(n^2).) But this can be easily avoided, if you re-thing the procedure to consume the list the other way around. The new structure of the function could look like this:
fromDigits' :: [Int] -> Integer
fromDigits' = f 0
where
f s [] = s
f s (x:xs) = f (s + ...) xs
After that, try using foldl to express f and finally replace it with Foldable.foldl.
You should avoid the use of length and write your function using foldl (or foldl'):
fromDigits :: [Int] -> Integer
fromDigits ds = foldl (\s d -> s*10 + (fromIntegral d)) 0 ds
From this a generalization to any Foldable should be clear.
A better way to solve this is to build up a list of your powers of 10. This is quite simple using iterate:
powersOf :: Num a => a -> [a]
powersOf n = iterate (*n) 1
Then you just need to multiply these powers of 10 by their respective values in the list of digits. This is easily accomplished with zipWith (*), but you have to make sure it's in the right order first. This basically just means that you should re-order your digits so that they're in descending order of magnitude instead of ascending:
zipWith (*) (powersOf 10) $ reverse xs
But we want it to return an Integer, not Int, so let's through a map fromIntegral in there
zipWith (*) (powersOf 10) $ map fromIntegral $ reverse xs
And all that's left is to sum them up
fromDigits :: [Int] -> Integer
fromDigits xs = sum $ zipWith (*) (powersOf 10) $ map fromIntegral $ reverse xs
Or for the point-free fans
fromDigits = sum . zipWith (*) (powersOf 10) . map fromIntegral . reverse
Now, you can also use a fold, which is basically just a pure for loop where the function is your loop body, the initial value is, well, the initial state, and the list you provide it is the values you're looping over. In this case, your state is a sum and what power you're on. We could make our own data type to represent this, or we could just use a tuple with the first element being the current total and the second element being the current power:
fromDigits xs = fst $ foldr go (0, 1) xs
where
go digit (s, power) = (s + digit * power, power * 10)
This is roughly equivalent to the Python code
def fromDigits(digits):
def go(digit, acc):
s, power = acc
return (s + digit * power, power * 10)
state = (0, 1)
for digit in digits:
state = go(digit, state)
return state[0]
Such a simple function can carry all its state in its bare arguments. Carry around an accumulator argument, and the operation becomes trivial.
fromDigits :: [Int] -> Integer
fromDigits xs = fromDigitsA xs 0 # 0 is the current accumulator value
fromDigitsA [] acc = acc
fromDigitsA (x:xs) acc = fromDigitsA xs (acc * 10 + toInteger x)
If you're really determined to use a right fold for this, you can combine calculating length xs with the calculation like this (taking the liberty of defining fromDigits [] = 0):
fromDigits xn = let (x, _) = fromDigits' xn in x where
fromDigits' [] = (0, 0)
fromDigits' (x:xn) = (toInteger x * 10 ^ l + y, l + 1) where
(y, l) = fromDigits' xn
Now it should be obvious that this is equivalent to
fromDigits xn = fst $ foldr (\ x (y, l) -> (toInteger x * 10^l + y, l + 1)) (0, 0) xn
The pattern of adding an extra component or result to your accumulator, and discarding it once the fold returns, is a very general one when you're re-writing recursive functions using folds.
Having said that, a foldr with a function that is always strict in its second parameter is a really, really bad idea (excessive stack usage, maybe a stack overflow on long lists) and you really should write fromDigits as a foldl as some of the other answers have suggested.
If you want to "fold with state", probably Traversable is the abstraction you're looking for. One of the methods defined in Traversable class is
traverse :: Applicative f => (a -> f b) -> t a -> f (t b)
Basically, traverse takes a "stateful function" of type a -> f b and applies it to every function in the container t a, resulting in a container f (t b). Here, f can be State, and you can use traverse with function of type Int -> State Integer (). It would build an useless data structure (list of units in your case), but you can just discard it. Here's a solution to your problem using Traversable:
import Control.Monad.State
import Data.Traversable
sumDigits :: Traversable t => t Int -> Integer
sumDigits cont = snd $ runState (traverse action cont) 0
where action x = modify ((+ (fromIntegral x)) . (* 10))
test1 = sumDigits [1, 4, 5, 6]
However, if you really don't like building discarded data structure, you can just use Foldable with somewhat tricky Monoid implementation: store not only computed result, but also 10^n, where n is count of digits converted to this value. This additional information gives you an ability to combine two values:
import Data.Foldable
import Data.Monoid
data Digits = Digits
{ value :: Integer
, power :: Integer
}
instance Monoid Digits where
mempty = Digits 0 1
(Digits d1 p1) `mappend` (Digits d2 p2) =
Digits (d1 * p2 + d2) (p1 * p2)
sumDigitsF :: Foldable f => f Int -> Integer
sumDigitsF cont = value $ foldMap (\x -> Digits (fromIntegral x) 10) cont
test2 = sumDigitsF [0, 4, 5, 0, 3]
I'd stick with first implementation. Although it builds unnecessary data structure, it's shorter and simpler to understand (as far as a reader understands Traversable).
I am trying to convert a float to a bit list.
Here is my idea:
read all bits info from a float to an Int64
mod 2 64 times and each time store the 0 or 1 to a list
Deal with the negative case (applying 2's complement plus one)
The code is like this:
let convert_non_neg_64 n =
let rec collect acc i j =
if Int64.compare j (Int64.of_int 64) <> -1 then acc
else collect ((Int64.rem i (Int64.of_int 2) |> Int64.to_int)::acc) (Int64.div i (Int64.of_int 2)) (Int64.succ j)
in
collect [] n Int64.zero
let negatify l =
let rl = List.rev_map (fun x -> if x = 0 then 1 else 0) l in
let rec plus1 extra acc = function
| [] -> acc
| hd::tl when extra = 0 -> plus1 0 (hd::acc) tl
| hd::tl -> if hd = 0 then plus1 0 (1::acc) tl else plus1 1 (0::acc) tl
in
plus1 1 [] rl
let convert64 n =
let nnl = convert_non_neg_64 (Int64.abs n) in
if Int64.compare n Int64.zero <> -1 then nnl
else negatify nnl
let bits_of_float fn = Int64.bits_of_float fn |> convert64
I think the code is fine.
My question is is there an easier way?
Also, all those Int64 operations are really ugly, any good way to simplify it?
This is what I came up with.
let bitlist_of_float f =
let rec blist b i64 =
if b >= 64 then
[]
else
let bit =
Int64.(if logand i64 (shift_left 1L b) = 0L then 0 else 1)
in
bit :: blist (b + 1) i64
in
blist 0 (Int64.bits_of_float f)
The head of the list is the least significant bit (which I believe is the right way to do things). I have not verified the answers, so there could be an error or two.
Update
I believe this code will give you the bits of your floating value. That's all I mean by "faithful copy." If you wanted to modify the bits somehow, then you'd need to work on them a little more. I'm not sure what you have in mind to do with them.