Zipf
Power-law probabilities over ranks — the signature of word frequencies and city sizes.
Reference / Complete index
Seventy-six distributions.
Search every distribution represented in the source chart. Fourteen foundational models have a dedicated interactive page; every entry links back into the relationship material.
Power-law probabilities over ranks — the signature of word frequencies and city sizes.
Equal chance on each of n outcomes: a fair die with any number of faces.
The discrete uniform under its classical name, flat across consecutive integers.
Binomial counts whose success rate is itself Beta-distributed — overdispersed by design.
Draws without replacement until a fixed number of failures arrives.
Zipf stretched to infinite support, normalized by the Riemann zeta function.
Heavy at one, thin in the tail — the count law behind species-abundance models.
The family umbrella: any count law whose probabilities come from a power-series expansion.
Counts independent arrivals in a fixed exposure; a sparse-event limit of the binomial.
The number of successes in n independent Bernoulli trials.
Successes in draws without replacement — quality control from a finite lot.
One uncertain yes/no outcome—the atom from which binomial counts are built.
Pascal waiting times whose success rate is itself Beta-distributed.
Poisson counts with a Gamma-random rate — the negative binomial in mixture form.
Trials until the first success; the discrete memoryless law.
Trials until the r-th success — the negative binomial counted from one.
Urn draws that reinforce themselves; contagion added to the binomial.
Weibull-shaped hazards on the integers, for lifetimes counted in cycles.
The central limit attractor and the default local model for additive noise.
The joint conjugate prior for a normal mean and precision together.
The parameter-free normal reached by standardizing X to (X−μ)/σ.
Defined by an arctangent CDF — a heavy-tailed relative of the Cauchy.
Bell-shaped like the normal but with heavier tails and closed-form everything.
A heavy-tailed ratio of normals; its mean and variance do not exist.
First-passage time of drifting Brownian motion; skewed, positive, tractable.
The inverse Gaussian with its mean pinned at one.
Chi-square shifted by a noncentrality term; the power side of tests.
The square root of a chi-square — lengths of Gaussian vectors.
Sum of squared standard normals; a gamma special case central to inference.
The logarithm of a gamma variate; extreme-value analysis on the log scale.
Reciprocal of a gamma — the conjugate prior for a normal variance.
Multiplicative noise made law: exponentiate a normal, get skewed positive data.
One shape knob away from exponential, Weibull, gamma, and lognormal at once.
Positive sums and waiting time to multiple arrivals; contains exponential and chi-square.
The Beta odds ratio p/(1−p); also known as the beta prime.
A flexible model for proportions and the conjugate prior for Bernoulli probability.
U-shaped on [0,1] — where random walks spend their time.
Gompertz aging plus a constant background hazard; the actuarial mortality law.
Beta with a noncentrality shift, powering F-test calculations.
Student t with one degree of freedom; no mean, no variance.
Normal uncertainty with an estimated scale; approaches normal as ν grows.
A chain of exponential stages with different rates — series systems.
Sum of k exponential waits; the gamma with integer shape, born in telephone traffic.
Exponentially accelerating hazard — the demographer’s law of mortality.
A two-parameter lifetime law that can bend its hazard both ways.
Waiting time to the next Poisson arrival; the continuous memoryless law.
Ratio of two scaled chi-squares; the workhorse of analysis of variance.
The t statistic under a false null; sample-size and power planning.
Noncentrality in both numerator and denominator of the t.
A mixture of exponentials, more variable than any single one.
A lifetime model whose hazard climbs steadily above the exponential baseline.
The generalized normal: one shape parameter sweeps from spiky to near-uniform.
Uniform on [0,1] — the raw material of every random-variate generator.
A lifetime law capable of bathtub-shaped hazards.
Beta-like shapes with closed-form CDF and quantiles, built for order statistics.
Power-function densities on the unit interval; Beta with one shape fixed.
The tent on [−1, 1]: the sum of two uniforms.
Back-to-back exponentials; the sparsity-inducing prior behind the lasso.
The 80/20 tail: polynomial decay above a threshold.
Increasing, decreasing, or bathtub hazards from a single three-parameter law.
Speeds of two-dimensional Gaussian vectors; Weibull with shape two.
The F under an alternative hypothesis — ANOVA power analysis.
Noncentrality in both chi-squares of the F ratio.
A lifetime model whose shape controls whether hazard rises, falls, or stays constant.
Logistic on the log scale; heavy-tailed lifetimes with closed-form quantiles.
The law of leading digits: 1 appears about thirty percent of the time.
Two-sided power: the triangular generalized with a peak-sharpness exponent.
A polynomial density rising toward the right end of its support.
Equal density over an interval and the source of inverse-transform simulation.
The normal of the circle: directional data with a concentration parameter.
Null law of the largest gap between empirical and true CDFs.
Min, mode, max — the expert-elicitation stand-in for scarce data.
Sigmoid CDF with slightly heavy tails; the link behind logistic regression.
Pareto shifted to start at zero; long tails for incomes and failures.
The tail law of exceedances over a threshold in extreme-value theory.
Limit law of maxima — Gumbel and friends for floods, records, and crashes.