class Crowbar::PRNG

• Crowbar::PRNG
• Reference
• Object

Overview

High-performance, deterministic 64-bit pseudo-random number generator. Uses Xoshiro256++ seeded via SplitMix64 for uniform statistical properties, long period (2^256 - 1), and complete reproducibility across platforms.

Defined in:

crowbar/prng.cr

Constructors

• .new(seed : UInt64 = PRNG.default_seed)

Class Method Summary

• .default_seed : UInt64

Instance Method Summary

• #choice(array : Array(T)) : T forall T

  Random element from array

• #choice?(array : Array(T)) : T | Nil forall T
• #next_u32 : UInt32

  Generate next 32-bit unsigned integer

• #next_u64 : UInt64

  Generate next raw 64-bit unsigned integer using Xoshiro256++

• #rand(max : Int) : Int32

  Random integer in [0, max)

• #rand(range : Range(Int32, Int32)) : Int32

  Random integer within inclusive range (e.g.

• #rand_bool(prob : Float64 = 0.5) : Bool

  Returns true with probability p (default 0.5)

• #rand_float : Float64

  Random Float64 in [0.0, 1.0)

• #rand_log(max_power : Int32 = 14) : Int32

  Log-scale random distribution (useful for repeats, burst lengths, and byte counts) Generates values from 1 up to 2^max_power with logarithmic bias towards smaller values

• #sample(array : Array(T), count : Int32) : Array(T) forall T

  Reservoir sample or simple sample

• #seed : UInt64
• #seed=(new_seed : UInt64)
• #seek(offset : Int) : Nil

  Fast-forwards PRNG state by offset iterations

• #shuffle(array : Array(T)) : Array(T) forall T

  Non-destructive shuffle

• #shuffle!(array : Array(T)) : Array(T) forall T

  Shuffle array in-place (Fisher-Yates) and return it

Constructor Detail

Class Method Detail

Instance Method Detail