SRL::Math::Random< T > Class Template Reference

Detailed Description

template<typename T>
class SRL::Math::Random< T >

High-performance pseudo-random number generator optimized for Saturn hardware.

The Random class implements a fast, lightweight Xorshift pseudo-random number generator (PRNG) that provides high-quality random numbers with excellent statistical properties while maintaining minimal computational overhead. This implementation is specifically optimized for Saturn hardware and game development scenarios where performance is critical.

Key features:

• Fast generation with minimal state (single variable)
• Long period relative to state size
• Good statistical distribution properties
• Deterministic sequence for reproducible results
• Configurable for different integral types
• No floating-point operations in core algorithm

Implementation details:

• Uses Xorshift algorithm (https://en.wikipedia.org/wiki/Xorshift)
• Shift parameters are optimized based on bit width of the template type
• Automatically adjusts algorithm for different integral types
• Period length depends on the bit width of type T

Statistical properties:

• Passes most common randomness tests (Die-Hard, TestU01)
• Even distribution across the entire range of type T
• Low correlation between consecutive outputs
• Suitable for most game development and simulation needs

Common applications:

• Procedural content generation
• AI decision making
• Particle effects
• Gameplay mechanics (damage ranges, critical hits)
• Simulation of natural phenomena

Performance considerations:

• Extremely fast (few CPU operations per number)
• Small memory footprint (single state variable)
• No branching in core algorithm
• No division operations

Note

While this PRNG is suitable for games and simulations, it is NOT cryptographically secure and should not be used for security-related applications or where unpredictability is critical.

Template Parameters

T	Integral type for random number generation (determines range and period)

See also

Fxp For fixed-point representation of random values

#include <random.hpp>

Public Member Functions
	Random (T seed)
	Construct a new pseudo-random number generator.
	~Random ()
	Destroy the pseudo-random number generator object.
T	GetNumber ()
	Get next pseudo-random number.
T	GetNumber (T from, T to)
	Get next pseudo-random number.