Control.Monad.Random.Lazy

A random monad parameterized by the type g of the generator to carry.

The return function leaves the generator unchanged, while >>= uses the final generator of the first computation as the initial generator of the second.

Construct a random monad computation from a function. (The inverse of runRand.)

Unwrap a random monad computation as a function. (The inverse of liftRand.)

Evaluate a random computation with the given initial generator and return the final value, discarding the final generator.

• evalRand m s = fst (runRand m s)

Evaluate a random computation with the given initial generator and return the final generator, discarding the final value.

• execRand m s = snd (runRand m s)

Map both the return value and final generator of a computation using the given function.

• runRand (mapRand f m) = f . runRand m

withRand f m executes action m on a generator modified by applying f.

• withRand f m = modify f >> m

Evaluate a random computation in the IO monad, splitting the global standard generator to get a new one for the computation.

A random transformer monad parameterized by:

• g - The generator.

• m - The inner monad.

The return function leaves the generator unchanged, while >>= uses the final generator of the first computation as the initial generator of the second.

Construct a random monad computation from an impure function. (The inverse of runRandT.)

Unwrap a random monad computation as an impure function. (The inverse of liftRandT.)

Evaluate a random computation with the given initial generator and return the final value, discarding the final generator.

• evalRandT m g = liftM fst (runRandT m g)

Evaluate a random computation with the given initial generator and return the final generator, discarding the final value.

• execRandT m g = liftM snd (runRandT m g)

Map both the return value and final generator of a computation using the given function.

• runRandT (mapRandT f m) = f . runRandT m

withRandT f m executes action m on a generator modified by applying f.

• withRandT f m = modify f >> m

Evaluate a random computation that is embedded in the IO monad, splitting the global standard generator to get a new one for the computation.

The class of types for which random values can be generated. Most instances of Random will produce values that are uniformly distributed on the full range, but for those types without a well-defined "full range" some sensible default subrange will be selected.

Random exists primarily for backwards compatibility with version 1.1 of this library. In new code, use the better specified Uniform and UniformRange instead.

A type class for data with a finite number of inhabitants. This type class is used in default implementations of System.Random.Stateful.Uniform.

Users are not supposed to write instances of Finite manually. There is a default implementation in terms of Generic instead.

>>> :set -XDeriveGeneric -XDeriveAnyClass>>> import GHC.Generics (Generic)>>> data MyBool = MyTrue | MyFalse deriving (Generic, Finite)>>> data Action = Code MyBool | Eat (Maybe Bool) | Sleep deriving (Generic, Finite)

RandomGen is an interface to pure pseudo-random number generators.

StdGen is the standard RandomGen instance provided by this library.

The standard pseudo-random number generator.

The class of types for which a uniformly distributed value can be drawn from all possible values of the type.

The class of types for which a uniformly distributed value can be drawn from a range.

Constructs a StdGen deterministically.

Generates a ByteString of the specified size using a pure pseudo-random number generator. See uniformByteStringM for the monadic version.

Examples

>>> import System.Random>>> import Data.ByteString>>> let pureGen = mkStdGen 137>>> unpack . fst . genByteString 10 $ pureGen[51,123,251,37,49,167,90,109,1,4]

Gets the global pseudo-random number generator. Extracts the contents of globalStdGen

Uses the supplied function to get a value from the current global random generator, and updates the global generator with the new generator returned by the function. For example, rollDice produces a pseudo-random integer between 1 and 6:

>>> rollDice = getStdRandom (randomR (1, 6))>>> replicateM 10 (rollDice :: IO Int)[5,6,6,1,1,6,4,2,4,1]

This is an outdated function and it is recommended to switch to its equivalent applyAtomicGen instead, possibly with the globalStdGen if relying on the global state is acceptable.

>>> import System.Random.Stateful>>> rollDice = applyAtomicGen (uniformR (1, 6)) globalStdGen>>> replicateM 10 (rollDice :: IO Int)[4,6,1,1,4,4,3,2,1,2]

Initialize StdGen using system entropy (i.e. /dev/urandom) when it is available, while falling back on using system time as the seed.

Applies split to the current global pseudo-random generator globalStdGen, updates it with one of the results, and returns the other.

A variant of randomM that uses the global pseudo-random number generator globalStdGen.

>>> import Data.Int>>> randomIO :: IO Int32-1580093805

This function is equivalent to getStdRandom random and is included in this interface for historical reasons and backwards compatibility. It is recommended to use uniformM instead, possibly with the globalStdGen if relying on the global state is acceptable.

>>> import System.Random.Stateful>>> uniformM globalStdGen :: IO Int32-1649127057

A variant of randomRM that uses the global pseudo-random number generator globalStdGen

>>> randomRIO (2020, 2100) :: IO Int2040

Similar to randomIO, this function is equivalent to getStdRandom randomR and is included in this interface for historical reasons and backwards compatibility. It is recommended to use uniformRM instead, possibly with the globalStdGen if relying on the global state is acceptable.

>>> import System.Random.Stateful>>> uniformRM (2020, 2100) globalStdGen :: IO Int2079

Sets the global pseudo-random number generator. Overwrites the contents of globalStdGen

The Monad class defines the basic operations over a monad, a concept from a branch of mathematics known as category theory. From the perspective of a Haskell programmer, however, it is best to think of a monad as an abstract datatype of actions. Haskell's do expressions provide a convenient syntax for writing monadic expressions.

Instances of Monad should satisfy the following:

Left identity

return a >>= k = k a

Right identity

m >>= return = m

Associativity

m >>= (\x -> k x >>= h) = (m >>= k) >>= h

Furthermore, the Monad and Applicative operations should relate as follows:

• pure = return

• m1 <*> m2 = m1 >>= (\x1 -> m2 >>= (\x2 -> return (x1 x2)))

The above laws imply:

• fmap f xs  =  xs >>= return . f

• (>>) = (*>)

and that pure and (<*>) satisfy the applicative functor laws.

The instances of Monad for GHC.List.List, Maybe and System.IO.IO defined in the Prelude satisfy these laws.

Promote a function to a monad. This is equivalent to fmap but specialised to Monads.

Monads that also support choice and failure.

A type f is a Functor if it provides a function fmap which, given any types a and b lets you apply any function from (a -> b) to turn an f a into an f b, preserving the structure of f. Furthermore f needs to adhere to the following:

Identity

fmap id == id

Composition

fmap (f . g) == fmap f . fmap g

Note, that the second law follows from the free theorem of the type fmap and the first law, so you need only check that the former condition holds. See these articles by School of Haskell or David Luposchainsky for an explanation.

Same as >>=, but with the arguments interchanged.

as >>= f == f =<< as

Map each element of a structure to a monadic action, evaluate these actions from left to right, and ignore the results. For a version that doesn't ignore the results see Data.Traversable.mapM.

mapM_ is just like traverse_, but specialised to monadic actions.

Evaluate each monadic action in the structure from left to right, and ignore the results. For a version that doesn't ignore the results see Data.Traversable.sequence.

sequence_ is just like sequenceA_, but specialised to monadic actions.

Map each element of a structure to a monadic action, evaluate these actions from left to right, and collect the results. For a version that ignores the results see Data.Foldable.mapM_.

Examples

mapM is literally a traverse with a type signature restricted to Monad. Its implementation may be more efficient due to additional power of Monad.

Evaluate each monadic action in the structure from left to right, and collect the results. For a version that ignores the results see Data.Foldable.sequence_.

Examples

Basic usage:

The first two examples are instances where the input and and output of sequence are isomorphic.

>>> sequence $ Right [1,2,3,4][Right 1,Right 2,Right 3,Right 4]

>>> sequence $ [Right 1,Right 2,Right 3,Right 4]Right [1,2,3,4]

The following examples demonstrate short circuit behavior for sequence.

>>> sequence $ Left [1,2,3,4]Left [1,2,3,4]

>>> sequence $ [Left 0, Right 1,Right 2,Right 3,Right 4]Left 0

The join function is the conventional monad join operator. It is used to remove one level of monadic structure, projecting its bound argument into the outer level.

'join bss' can be understood as the do expression

do bs <- bss
   bs

Examples

>>> join [[1, 2, 3], [4, 5, 6], [7, 8, 9]][1,2,3,4,5,6,7,8,9]

>>> join (Just (Just 3))Just 3

A common use of join is to run an IO computation returned from an GHC.Conc.STM transaction, since GHC.Conc.STM transactions can't perform IO directly. Recall that

GHC.Internal.Conc.atomically :: STM a -> IO a

is used to run GHC.Conc.STM transactions atomically. So, by specializing the types of GHC.Internal.Conc.atomically and join to

GHC.Internal.Conc.atomically :: STM (IO b) -> IO (IO b)
join       :: IO (IO b)  -> IO b

we can compose them as

join . GHC.Internal.Conc.atomically :: STM (IO b) -> IO b

to run an GHC.Conc.STM transaction and the IO action it returns.

In many situations, the liftM operations can be replaced by uses of ap, which promotes function application.

return f `ap` x1 `ap` ... `ap` xn

is equivalent to

liftM<n> f x1 x2 ... xn

Examples

>>> pure (\x y z -> x + y * z) `ap` Just 1 `ap` Just 5 `ap` Just 10Just 51

Promote a function to a monad, scanning the monadic arguments from left to right.

Examples

>>> liftM2 (+) [0,1] [0,2][0,2,1,3]

>>> liftM2 (+) (Just 1) NothingNothing

>>> liftM2 (+) (+ 3) (* 2) 518

Promote a function to a monad, scanning the monadic arguments from left to right (cf. liftM2).

Promote a function to a monad, scanning the monadic arguments from left to right (cf. liftM2).

Promote a function to a monad, scanning the monadic arguments from left to right (cf. liftM2).

Conditional execution of Applicative expressions. For example,

Examples

when debug (putStrLn "Debugging")

will output the string Debugging if the Boolean value debug is True, and otherwise do nothing.

>>> putStr "pi:" >> when False (print 3.14159)pi:

Strict version of Data.Functor.<$>.

Right-to-left composition of Kleisli arrows. (>=>), with the arguments flipped.

Note how this operator resembles function composition (.):

(.)   ::            (b ->   c) -> (a ->   b) -> a ->   c
(<=<) :: Monad m => (b -> m c) -> (a -> m b) -> a -> m c

Left-to-right composition of Kleisli arrows.

'(bs >=> cs) a' can be understood as the do expression

do b <- bs a
   cs b

or in terms of (>>=) as

bs a >>= cs

This generalizes the list-based filter function.

runIdentity (filterM (Identity . p) xs) == filter p xs

Examples

>>> filterM (\x -> do      putStrLn ("Keep: " ++ show x ++ "?")      answer <- getLine      pure (answer == "y"))    [1, 2, 3]Keep: 1?yKeep: 2?nKeep: 3?y[1,3]

>>> filterM (\x -> do      putStr (show x)      x' <- readLn      pure (x == x'))    [1, 2, 3]122233[2,3]

The foldM function is analogous to foldl, except that its result is encapsulated in a monad. Note that foldM works from left-to-right over the list arguments. This could be an issue where (>>) and the `folded function' are not commutative.

foldM f a1 [x1, x2, ..., xm]

==

do
  a2 <- f a1 x1
  a3 <- f a2 x2
  ...
  f am xm

If right-to-left evaluation is required, the input list should be reversed.

Note: foldM is the same as foldlM

Like foldM, but discards the result.

Repeat an action indefinitely.

Examples

A common use of forever is to process input from network sockets, System.IO.Handles, and channels (e.g. Control.Concurrent.MVar.MVar and Chan).

For example, here is how we might implement an echo server, using forever both to listen for client connections on a network socket and to echo client input on client connection handles:

echoServer :: Socket -> IO ()
echoServer socket = forever $ do
  client <- accept socket
  forkFinally (echo client) (\_ -> hClose client)
  where
    echo :: Handle -> IO ()
    echo client = forever $
      hGetLine client >>= hPutStrLn client

Note that "forever" isn't necessarily non-terminating. If the action is in a MonadPlus and short-circuits after some number of iterations. then forever actually returns mzero, effectively short-circuiting its caller.

The mapAndUnzipM function maps its first argument over a list, returning the result as a pair of lists. This function is mainly used with complicated data structures or a state monad.

Direct MonadPlus equivalent of filter.

Examples

The filter function is just mfilter specialized to the list monad:

filter = ( mfilter :: (a -> Bool) -> [a] -> [a] )

An example using mfilter with the Maybe monad:

>>> mfilter odd (Just 1)Just 1>>> mfilter odd (Just 2)Nothing

replicateM n act performs the action act n times, and then returns the list of results.

replicateM n (pure x) == replicate n x

Examples

>>> replicateM 3 getLinehiheyahiya["hi","heya","hiya"]

>>> import Control.Monad.State>>> runState (replicateM 3 $ state $ \s -> (s, s + 1)) 1([1,2,3],4)

Like replicateM, but discards the result.

Examples

>>> replicateM_ 3 (putStr "a")aaa

The reverse of when.

Examples

>>> do x <- getLine       unless (x == "hi") (putStrLn "hi!")comingupwithexamplesisdifficulthi!

>>> unless (pi > exp 1) NothingJust ()

The zipWithM function generalizes zipWith to arbitrary applicative functors.

zipWithM_ is the extension of zipWithM which ignores the final result.

forM_ is mapM_ with its arguments flipped. For a version that doesn't ignore the results see Data.Traversable.forM.

forM_ is just like for_, but specialised to monadic actions.

The sum of a collection of actions using (<|>), generalizing concat.

msum is just like asum, but specialised to MonadPlus.

Examples

Basic usage, using the MonadPlus instance for Maybe:

>>> msum [Just "Hello", Nothing, Just "World"]Just "Hello"

void value discards or ignores the result of evaluation, such as the return value of an System.IO.IO action.

Examples

Replace the contents of a Maybe Int with unit:

>>> void NothingNothing

>>> void (Just 3)Just ()

Replace the contents of an Either Int Int with unit, resulting in an Either Int ():

>>> void (Left 8675309)Left 8675309

>>> void (Right 8675309)Right ()

Replace every element of a list with unit:

>>> void [1,2,3][(),(),()]

Replace the second element of a pair with unit:

>>> void (1,2)(1,())

Discard the result of an System.IO.IO action:

>>> mapM print [1,2]12[(),()]

>>> void $ mapM print [1,2]12

forM is mapM with its arguments flipped. For a version that ignores the results see Data.Foldable.forM_.

Conditional failure of Alternative computations. Defined by

guard True  = pure ()
guard False = empty

Examples

Common uses of guard include conditionally signalling an error in an error monad and conditionally rejecting the current choice in an Alternative-based parser.

As an example of signalling an error in the error monad Maybe, consider a safe division function safeDiv x y that returns Nothing when the denominator y is zero and Just (x `div` y) otherwise. For example:

>>> safeDiv 4 0Nothing

>>> safeDiv 4 2Just 2

A definition of safeDiv using guards, but not guard:

safeDiv :: Int -> Int -> Maybe Int
safeDiv x y | y /= 0    = Just (x `div` y)
            | otherwise = Nothing

A definition of safeDiv using guard and Monad do-notation:

safeDiv :: Int -> Int -> Maybe Int
safeDiv x y = do
  guard (y /= 0)
  return (x `div` y)

When a value is bound in do-notation, the pattern on the left hand side of <- might not match. In this case, this class provides a function to recover.

A Monad without a MonadFail instance may only be used in conjunction with pattern that always match, such as newtypes, tuples, data types with only a single data constructor, and irrefutable patterns (~pat).

Instances of MonadFail should satisfy the following law: fail s should be a left zero for >>=,

fail s >>= f  =  fail s

If your Monad is also MonadPlus, a popular definition is

fail _ = mzero

fail s should be an action that runs in the monad itself, not an exception (except in instances of MonadIO). In particular, fail should not be implemented in terms of error.

Monads having fixed points with a 'knot-tying' semantics. Instances of MonadFix should satisfy the following laws:

Purity

mfix (return . h) = return (fix h)

Left shrinking (or Tightening)

mfix (\x -> a >>= \y -> f x y) = a >>= \y -> mfix (\x -> f x y)

Sliding

mfix (liftM h . f) = liftM h (mfix (f . h))

, for strict

h

.

Nesting

mfix (\x -> mfix (\y -> f x y)) = mfix (\x -> f x x)

This class is used in the translation of the recursive do notation supported by GHC and Hugs.

fix f is the least fixed point of the function f, i.e. the least defined x such that f x = x.

When f is strict, this means that because, by the definition of strictness, f ⊥ = ⊥ and such the least defined fixed point of any strict function is ⊥.

Examples

We can write the factorial function using direct recursion as

>>> let fac n = if n <= 1 then 1 else n * fac (n-1) in fac 5120

This uses the fact that Haskell’s let introduces recursive bindings. We can rewrite this definition using fix,

Instead of making a recursive call, we introduce a dummy parameter rec; when used within fix, this parameter then refers to fix’s argument, hence the recursion is reintroduced.

>>> fix (\rec n -> if n <= 1 then 1 else n * rec (n-1)) 5120

Using fix, we can implement versions of repeat as fix . (:) and cycle as fix . (++)

>>> take 10 $ fix (0:)[0,0,0,0,0,0,0,0,0,0]

>>> map (fix (\rec n -> if n < 2 then n else rec (n - 1) + rec (n - 2))) [1..10][1,1,2,3,5,8,13,21,34,55]

Implementation Details

The current implementation of fix uses structural sharing

fix f = let x = f x in x

A more straightforward but non-sharing version would look like

fix f = f (fix f)

Monads in which IO computations may be embedded. Any monad built by applying a sequence of monad transformers to the IO monad will be an instance of this class.

Instances should satisfy the following laws, which state that liftIO is a transformer of monads:

• liftIO . return = return

• liftIO (m >>= f) = liftIO m >>= (liftIO . f)

The class of monad transformers. For any monad m, the result t m should also be a monad, and lift should be a monad transformation from m to t m, i.e. it should satisfy the following laws:

• lift . return = return

• lift (m >>= f) = lift m >>= (lift . f)

Since 0.6.0.0 and for GHC 8.6 and later, the requirement that t m be a Monad is enforced by the implication constraint forall m. Monad m => Monad (t m) enabled by the QuantifiedConstraints extension.

Ambiguity error with GHC 9.0 to 9.2.2

These versions of GHC have a bug (https://gitlab.haskell.org/ghc/ghc/-/issues/20582) which causes constraints like

(MonadTrans t, forall m. Monad m => Monad (t m)) => ...

to be reported as ambiguous. For transformers 0.6 and later, this can be fixed by removing the second constraint, which is implied by the first.