Module distribution 

Distribution analysis.

Empirical distribution functions, histogram binning, QQ-plot data, distribution fitting/testing, and kernel density estimation.

Examples

use u_analytics::distribution::{ecdf, histogram_bins, BinMethod};

let data = [1.0, 2.0, 3.0, 4.0, 5.0];
let (values, probs) = ecdf(&data).unwrap();
assert_eq!(values.len(), 5);
assert!((probs[4] - 1.0).abs() < 1e-10);

let bins = histogram_bins(&data, BinMethod::Sturges).unwrap();
assert!(bins.n_bins >= 2);

Structs

FitResult

Result of a distribution fit via maximum likelihood estimation.

HistogramBins

Result of histogram bin computation.

KdeResult

Result of kernel density estimation.

Enums

BandwidthMethod

Bandwidth selection method for kernel density estimation.

BinMethod

Method for computing optimal number of histogram bins.

Functions

ecdf

A point on the empirical CDF.

fit_best

Fits multiple distributions and returns results sorted by AIC (best first).

fit_beta

Fits a Beta distribution Beta(α, β) via MLE.

fit_exponential

Fits an Exponential distribution Exp(λ) via MLE.

fit_gamma

Fits a Gamma distribution Gamma(α, β) via MLE.

fit_lognormal

Fits a LogNormal distribution via MLE.

fit_normal

Fits a Normal distribution N(μ, σ²) via MLE.

fit_poisson

Fits a Poisson distribution via MLE.

histogram_bins

Computes optimal histogram bins using the specified method.

kde

Gaussian kernel density estimation.

kde_bandwidth

Computes the bandwidth for KDE using the specified method.

kde_evaluate

Evaluates the kernel density estimate at a single point.

ks_test_normal

Kolmogorov-Smirnov one-sample test against the standard normal distribution.

qq_plot_normal

Generates QQ-plot data (theoretical quantiles vs sample quantiles).