README

This library is used to pick random values from a set of registered values, where values with a higher weight have a larger probability to be picked.

Installation

To install this library using Composer:

composer require mschandr/weighted-random

📚 Documentation

• API Reference - Complete API documentation for all classes and methods
• CHANGELOG - Version history and migration guides
• Contributing Guide - Guidelines for contributing to the project

🚀 Usage

Basic Usage

use mschandr\WeightedRandom\WeightedRandom;

// Create a generator
$gen = WeightedRandom::createFloat();

// Register values with weights
$gen->registerValue('common', 7.0)
    ->registerValue('uncommon', 2.5)
    ->registerValue('rare', 0.5);

// Generate a random value
$result = $gen->generate();

// Generate multiple values
$results = $gen->generateMultiple(10);

// Generate unique values (no duplicates)
$unique = $gen->generateMultipleWithoutDuplicates(3);

Batch Registration

$gen->registerValues([
    'apple'  => 3.0,
    'banana' => 2.0,
    'cherry' => 1.0,
]);

Groups (Multiple Values, Single Weight)

// Register a group of values that share a single weight
$gen->registerGroup(['bronze', 'silver', 'gold'], 5.0);
// When the group is selected, one member is chosen uniformly at random

Fair Distribution (Bag System)

$bag = WeightedRandom::createBag();
$bag->registerValues(['rare' => 1, 'common' => 9]);

// Over 10 draws: exactly 1 rare, 9 common (then bag reshuffles)
$results = $bag->generateMultiple(10);

Distribution Introspection

// Get probability distribution
$distribution = $gen->getDistribution();
// Returns: ['apple' => 0.5, 'banana' => 0.333, 'cherry' => 0.167]

// Get probability of specific value
$prob = $gen->getProbability('apple'); // 0.5

// Calculate Shannon entropy (distribution randomness)
$entropy = $gen->getEntropy();

// For numeric values - statistical analysis
$gen->registerValues([1 => 1.0, 2 => 2.0, 3 => 1.0]);
$mean = $gen->getExpectedValue();      // Weighted mean
$variance = $gen->getVariance();       // Weighted variance
$stdDev = $gen->getStandardDeviation(); // Standard deviation

Decay/Boost (Dynamic Weight Adjustment)

// Manual weight adjustment
$gen->decayWeight('common', 0.8);  // Reduce weight to 80%
$gen->boostWeight('rare', 1.5);    // Increase weight by 50%

// Adjust all weights
$gen->decayAllWeights(0.9);  // Reduce all weights to 90%
$gen->boostAllWeights(1.2);  // Increase all weights by 20%

// Automatic adjustment based on selection frequency
$gen->enableSelectionTracking();

// Generate some values...
$gen->generateMultiple(100);

// Auto-adjust: frequently selected values get decayed, rare ones get boosted
$gen->autoAdjustWeights(0.5); // 0.5 = adjustment strength

// View selection counts
$counts = $gen->getSelectionCounts();

// Reset tracking
$gen->resetSelectionCounts();

Composite Generators (Nested/Hierarchical)

// Create a hierarchy of generators
$rareLoot = WeightedRandom::createFloat();
$rareLoot->registerValues(['legendary_sword' => 1.0, 'magic_ring' => 1.0]);

$commonLoot = WeightedRandom::createFloat();
$commonLoot->registerValues(['wooden_sword' => 3.0, 'bread' => 2.0]);

$lootBox = WeightedRandom::createFloat();
$lootBox->registerValue($rareLoot, 0.1);    // 10% chance of rare loot table
$lootBox->registerValue($commonLoot, 0.9);  // 90% chance of common loot table

$item = $lootBox->generate(); // Draws from nested generator

Requirements

• PHP 8.1 – 8.4
• Composer Seeded RNG requires PHP 8.2+. On PHP 8.1, those tests are automatically skipped.

🛠 Development

vendor/bin/phpunit -c phpunit.xml --color

GitHub Actions CI runs tests against PHP 8.1, 8.2, 8.3, 8.4.

License

MIT License.

Migration Guide (2.x → 3.x)

WeightedRandom 2.x introduces new features and stricter validation. If you’re upgrading from 1.x, here’s what you need to know:

What’s New

• Float weight support → Use 0.7 vs 0.3 without scaling up to integers.
• Seeded RNG with namespace isolation (PHP ≥ 8.2) → Deterministic, reproducible draws.
• Chaining API for cleaner code.
• Probability helpers:
  • normalizeWeights() → normalized distribution.
  • getProbability($value) → single-value probability.
• Bag System (v2.2+) → fairness via without-replacement draws.
• Stricter validation → safer, more predictable behavior.

Roadmap

1. Floats + Normalization
2. Validation Enhancements
3. Chaining API
4. Groups
5. Seeded RNG
6. Distribution Introspection
7. Bag System
8. Decay/Boost
9. Composite Generators
10. Code coverage and proper testing