SwiftStats

Pre-1.0 Notice: This library is under active development. The API may change between minor versions until 1.0.

A comprehensive statistical analysis library for Swift 6, featuring descriptive statistics, probability distributions, correlation analysis, and regression modeling.

Features

Descriptive Statistics

• Mean, median, mode
• Variance and standard deviation (population and sample)
• Minimum, maximum, range, sum, count
• Percentiles and quartiles
• Interquartile range (IQR)
• Frequency distributions and histograms

Probability Distributions

• Normal (Gaussian) Distribution
• Uniform Distribution
• Exponential Distribution
• Poisson Distribution

Each distribution includes:

• Probability Density Function (PDF)
• Cumulative Distribution Function (CDF)
• Random sampling
• Mean and variance

Correlation Analysis

• Pearson correlation coefficient
• Spearman rank correlation coefficient
• Covariance (population and sample)

Regression Analysis

• Simple linear regression
• Polynomial regression (arbitrary degree)
• R-squared coefficient of determination
• Residual analysis
• Mean squared error (MSE) and root mean squared error (RMSE)

Random Number Generation

• Seedable PRNG for reproducible results
• Normal distribution sampling (Box-Muller transform)
• Exponential distribution sampling
• Uniform distribution sampling

Requirements

• Swift 6.0+
• iOS 16.0+
• macOS 13.0+
• tvOS 16.0+
• watchOS 9.0+
• visionOS 1.0+

Installation

Swift Package Manager

Add the following to your Package.swift file:

dependencies: [
    .package(url: "https://github.com/CorvidLabs/swift-stats.git", from: "0.1.0")
]

Then add the target dependency:

.target(
    name: "YourTarget",
    dependencies: [
        .product(name: "Stats", package: "swift-stats")
    ]
)

Usage

Descriptive Statistics

import Stats

let values = [1.0, 2.0, 3.0, 4.0, 5.0]

// Calculate individual statistics
let mean = try values.mean()
let median = try values.median()
let stdDev = try values.standardDeviation()

// Calculate all statistics at once
let stats = try Statistics.calculate(from: values)
print("Mean: \(stats.mean)")
print("Median: \(stats.median)")
print("Standard Deviation: \(stats.standardDeviation)")

Percentiles and Quartiles

let data = Array(1...100).map { Double($0) }

// Calculate specific percentile
let p75 = try data.percentile(75)

// Calculate quartiles
let quartiles = try data.quartiles()
print("Q1: \(quartiles.q1), Q2: \(quartiles.q2), Q3: \(quartiles.q3)")

// Calculate interquartile range
let iqr = try data.interquartileRange()

Histograms

let values = Array(1...100).map { Double($0) }

// Create histogram with 10 bins
let histogram = try values.histogram(binCount: 10)
for bin in histogram.bins {
    print("[\(bin.lowerBound)-\(bin.upperBound)]: \(bin.frequency)")
}

// Create histogram with custom bin edges
let customHist = try values.histogram(binEdges: [0, 25, 50, 75, 100])

Probability Distributions

// Normal distribution
let normal = try NormalDistribution(mean: 100, standardDeviation: 15)
let pdf = normal.pdf(115)
let cdf = normal.cdf(115)
let samples = normal.sample(count: 1000)

// Uniform distribution
let uniform = try UniformDistribution(lowerBound: 0, upperBound: 10)
let randomValue = uniform.sample()

// Exponential distribution
let exponential = try ExponentialDistribution(lambda: 0.5)
let expSample = exponential.sample()

// Poisson distribution
let poisson = try PoissonDistribution(lambda: 5.0)
let probability = poisson.pmf(3) // P(X = 3)

Correlation Analysis

let x = [1.0, 2.0, 3.0, 4.0, 5.0]
let y = [2.0, 4.0, 6.0, 8.0, 10.0]

// Pearson correlation
let pearson = try x.pearsonCorrelation(with: y)

// Spearman correlation
let spearman = try x.spearmanCorrelation(with: y)

// Covariance
let covariance = try x.covariance(with: y)

Linear Regression

let x = [1.0, 2.0, 3.0, 4.0, 5.0]
let y = [2.0, 4.0, 6.0, 8.0, 10.0]

let regression = try LinearRegression.fit(x: x, y: y)
print("Slope: \(regression.slope)")
print("Intercept: \(regression.intercept)")
print("R²: \(regression.rSquared)")

// Make predictions
let predicted = regression.predict(6.0)

// Calculate residuals
let residuals = regression.residuals(x: x, y: y)

Polynomial Regression

let x = [1.0, 2.0, 3.0, 4.0, 5.0]
let y = [1.0, 4.0, 9.0, 16.0, 25.0] // y = x²

let regression = try PolynomialRegression.fit(x: x, y: y, degree: 2)
print("Coefficients: \(regression.coefficients)")
print("R²: \(regression.rSquared)")

// Make predictions
let predicted = regression.predict(6.0)

Reproducible Random Sampling

// Use a seeded random source for reproducibility
let randomSource = RandomSource(seed: 42)

let normal = try NormalDistribution(
    mean: 0,
    standardDeviation: 1,
    randomSource: randomSource
)

let samples = normal.sample(count: 100)

Architecture

The library is organized into focused modules:

• Descriptive: Statistical measures and frequency distributions
• Distributions: Probability distributions with PDF, CDF, and sampling
• Correlation: Correlation coefficients and covariance
• Regression: Linear and polynomial regression models
• Random: Seedable pseudo-random number generation

All types are Sendable and designed for Swift 6 strict concurrency.

Testing

The package includes comprehensive tests with known statistical values:

swift test

All 75 tests verify:

• Descriptive statistics calculations
• Distribution properties (PDF, CDF, mean, variance)
• Correlation coefficients
• Regression model accuracy
• Random number generation reproducibility

License

MIT License

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.