random is:exact

Pseudo-random number generation This package provides basic pseudo-random number generation, including the ability to split random number generators.

System.Random: pure pseudo-random number interface

In pure code, use System.Random.uniform and System.Random.uniformR from System.Random to generate pseudo-random numbers with a pure pseudo-random number generator like System.Random.StdGen. As an example, here is how you can simulate rolls of a six-sided die using System.Random.uniformR:
>>> let roll = uniformR (1, 6)        :: RandomGen g => g -> (Word, g)

>>> let rolls = unfoldr (Just . roll) :: RandomGen g => g -> [Word]

>>> let pureGen = mkStdGen 42

>>> take 10 (rolls pureGen)           :: [Word]
[1,1,3,2,4,5,3,4,6,2]
See System.Random for more details.

System.Random.Stateful: monadic pseudo-random number interface

In monadic code, use System.Random.Stateful.uniformM and System.Random.Stateful.uniformRM from System.Random.Stateful to generate pseudo-random numbers with a monadic pseudo-random number generator, or using a monadic adapter. As an example, here is how you can simulate rolls of a six-sided die using System.Random.Stateful.uniformRM:
>>> let rollM = uniformRM (1, 6)                 :: StatefulGen g m => g -> m Word

>>> let pureGen = mkStdGen 42

>>> runStateGen_ pureGen (replicateM 10 . rollM) :: [Word]
[1,1,3,2,4,5,3,4,6,2]
The monadic adapter System.Random.Stateful.runStateGen_ is used here to lift the pure pseudo-random number generator pureGen into the System.Random.Stateful.StatefulGen context. The monadic interface can also be used with existing monadic pseudo-random number generators. In this example, we use the one provided in the mwc-random package:
>>> import System.Random.MWC as MWC

>>> let rollM = uniformRM (1, 6)       :: StatefulGen g m => g -> m Word

>>> monadicGen <- MWC.create

>>> replicateM 10 (rollM monadicGen) :: IO [Word]
[2,3,6,6,4,4,3,1,5,4]
See System.Random.Stateful for more details.
The same as randomR, but using a default range determined by the type:
  • For bounded types (instances of Bounded, such as Char), the range is normally the whole type.
  • For floating point types, the range is normally the closed interval [0,1].
  • For Integer, the range is (arbitrarily) the range of Int.
The same as randomR, but using a default range determined by the type:
  • For bounded types (instances of Bounded, such as Char), the range is normally the whole type.
  • For floating point types, the range is normally the closed interval [0,1].
  • For Integer, the range is (arbitrarily) the range of Int.
Create a random Seed using an effectful source of randomness.
Pick a random element, using reservoir sampling
Pick a random element of the list.
Set random playing.
Toggle random mode.
Generate a random bytestring of length n. The PRNG is seeded from the system randomness source.
ioProperty $ ((fromIntegral n ===) . B.length) <$> random n
n > 4 ==> ioProperty $ (/=) <$> random n <*> random n
Generate a single random UUID.
Return hosts in random order.
Draw from a uniform distribution.