359 lines
9.1 KiB
JavaScript
359 lines
9.1 KiB
JavaScript
// https://d3js.org/d3-random/ v3.0.1 Copyright 2010-2021 Mike Bostock
|
|
(function (global, factory) {
|
|
typeof exports === 'object' && typeof module !== 'undefined' ? factory(exports) :
|
|
typeof define === 'function' && define.amd ? define(['exports'], factory) :
|
|
(global = typeof globalThis !== 'undefined' ? globalThis : global || self, factory(global.d3 = global.d3 || {}));
|
|
}(this, (function (exports) { 'use strict';
|
|
|
|
var defaultSource = Math.random;
|
|
|
|
var uniform = (function sourceRandomUniform(source) {
|
|
function randomUniform(min, max) {
|
|
min = min == null ? 0 : +min;
|
|
max = max == null ? 1 : +max;
|
|
if (arguments.length === 1) max = min, min = 0;
|
|
else max -= min;
|
|
return function() {
|
|
return source() * max + min;
|
|
};
|
|
}
|
|
|
|
randomUniform.source = sourceRandomUniform;
|
|
|
|
return randomUniform;
|
|
})(defaultSource);
|
|
|
|
var int = (function sourceRandomInt(source) {
|
|
function randomInt(min, max) {
|
|
if (arguments.length < 2) max = min, min = 0;
|
|
min = Math.floor(min);
|
|
max = Math.floor(max) - min;
|
|
return function() {
|
|
return Math.floor(source() * max + min);
|
|
};
|
|
}
|
|
|
|
randomInt.source = sourceRandomInt;
|
|
|
|
return randomInt;
|
|
})(defaultSource);
|
|
|
|
var normal = (function sourceRandomNormal(source) {
|
|
function randomNormal(mu, sigma) {
|
|
var x, r;
|
|
mu = mu == null ? 0 : +mu;
|
|
sigma = sigma == null ? 1 : +sigma;
|
|
return function() {
|
|
var y;
|
|
|
|
// If available, use the second previously-generated uniform random.
|
|
if (x != null) y = x, x = null;
|
|
|
|
// Otherwise, generate a new x and y.
|
|
else do {
|
|
x = source() * 2 - 1;
|
|
y = source() * 2 - 1;
|
|
r = x * x + y * y;
|
|
} while (!r || r > 1);
|
|
|
|
return mu + sigma * y * Math.sqrt(-2 * Math.log(r) / r);
|
|
};
|
|
}
|
|
|
|
randomNormal.source = sourceRandomNormal;
|
|
|
|
return randomNormal;
|
|
})(defaultSource);
|
|
|
|
var logNormal = (function sourceRandomLogNormal(source) {
|
|
var N = normal.source(source);
|
|
|
|
function randomLogNormal() {
|
|
var randomNormal = N.apply(this, arguments);
|
|
return function() {
|
|
return Math.exp(randomNormal());
|
|
};
|
|
}
|
|
|
|
randomLogNormal.source = sourceRandomLogNormal;
|
|
|
|
return randomLogNormal;
|
|
})(defaultSource);
|
|
|
|
var irwinHall = (function sourceRandomIrwinHall(source) {
|
|
function randomIrwinHall(n) {
|
|
if ((n = +n) <= 0) return () => 0;
|
|
return function() {
|
|
for (var sum = 0, i = n; i > 1; --i) sum += source();
|
|
return sum + i * source();
|
|
};
|
|
}
|
|
|
|
randomIrwinHall.source = sourceRandomIrwinHall;
|
|
|
|
return randomIrwinHall;
|
|
})(defaultSource);
|
|
|
|
var bates = (function sourceRandomBates(source) {
|
|
var I = irwinHall.source(source);
|
|
|
|
function randomBates(n) {
|
|
// use limiting distribution at n === 0
|
|
if ((n = +n) === 0) return source;
|
|
var randomIrwinHall = I(n);
|
|
return function() {
|
|
return randomIrwinHall() / n;
|
|
};
|
|
}
|
|
|
|
randomBates.source = sourceRandomBates;
|
|
|
|
return randomBates;
|
|
})(defaultSource);
|
|
|
|
var exponential = (function sourceRandomExponential(source) {
|
|
function randomExponential(lambda) {
|
|
return function() {
|
|
return -Math.log1p(-source()) / lambda;
|
|
};
|
|
}
|
|
|
|
randomExponential.source = sourceRandomExponential;
|
|
|
|
return randomExponential;
|
|
})(defaultSource);
|
|
|
|
var pareto = (function sourceRandomPareto(source) {
|
|
function randomPareto(alpha) {
|
|
if ((alpha = +alpha) < 0) throw new RangeError("invalid alpha");
|
|
alpha = 1 / -alpha;
|
|
return function() {
|
|
return Math.pow(1 - source(), alpha);
|
|
};
|
|
}
|
|
|
|
randomPareto.source = sourceRandomPareto;
|
|
|
|
return randomPareto;
|
|
})(defaultSource);
|
|
|
|
var bernoulli = (function sourceRandomBernoulli(source) {
|
|
function randomBernoulli(p) {
|
|
if ((p = +p) < 0 || p > 1) throw new RangeError("invalid p");
|
|
return function() {
|
|
return Math.floor(source() + p);
|
|
};
|
|
}
|
|
|
|
randomBernoulli.source = sourceRandomBernoulli;
|
|
|
|
return randomBernoulli;
|
|
})(defaultSource);
|
|
|
|
var geometric = (function sourceRandomGeometric(source) {
|
|
function randomGeometric(p) {
|
|
if ((p = +p) < 0 || p > 1) throw new RangeError("invalid p");
|
|
if (p === 0) return () => Infinity;
|
|
if (p === 1) return () => 1;
|
|
p = Math.log1p(-p);
|
|
return function() {
|
|
return 1 + Math.floor(Math.log1p(-source()) / p);
|
|
};
|
|
}
|
|
|
|
randomGeometric.source = sourceRandomGeometric;
|
|
|
|
return randomGeometric;
|
|
})(defaultSource);
|
|
|
|
var gamma = (function sourceRandomGamma(source) {
|
|
var randomNormal = normal.source(source)();
|
|
|
|
function randomGamma(k, theta) {
|
|
if ((k = +k) < 0) throw new RangeError("invalid k");
|
|
// degenerate distribution if k === 0
|
|
if (k === 0) return () => 0;
|
|
theta = theta == null ? 1 : +theta;
|
|
// exponential distribution if k === 1
|
|
if (k === 1) return () => -Math.log1p(-source()) * theta;
|
|
|
|
var d = (k < 1 ? k + 1 : k) - 1 / 3,
|
|
c = 1 / (3 * Math.sqrt(d)),
|
|
multiplier = k < 1 ? () => Math.pow(source(), 1 / k) : () => 1;
|
|
return function() {
|
|
do {
|
|
do {
|
|
var x = randomNormal(),
|
|
v = 1 + c * x;
|
|
} while (v <= 0);
|
|
v *= v * v;
|
|
var u = 1 - source();
|
|
} while (u >= 1 - 0.0331 * x * x * x * x && Math.log(u) >= 0.5 * x * x + d * (1 - v + Math.log(v)));
|
|
return d * v * multiplier() * theta;
|
|
};
|
|
}
|
|
|
|
randomGamma.source = sourceRandomGamma;
|
|
|
|
return randomGamma;
|
|
})(defaultSource);
|
|
|
|
var beta = (function sourceRandomBeta(source) {
|
|
var G = gamma.source(source);
|
|
|
|
function randomBeta(alpha, beta) {
|
|
var X = G(alpha),
|
|
Y = G(beta);
|
|
return function() {
|
|
var x = X();
|
|
return x === 0 ? 0 : x / (x + Y());
|
|
};
|
|
}
|
|
|
|
randomBeta.source = sourceRandomBeta;
|
|
|
|
return randomBeta;
|
|
})(defaultSource);
|
|
|
|
var binomial = (function sourceRandomBinomial(source) {
|
|
var G = geometric.source(source),
|
|
B = beta.source(source);
|
|
|
|
function randomBinomial(n, p) {
|
|
n = +n;
|
|
if ((p = +p) >= 1) return () => n;
|
|
if (p <= 0) return () => 0;
|
|
return function() {
|
|
var acc = 0, nn = n, pp = p;
|
|
while (nn * pp > 16 && nn * (1 - pp) > 16) {
|
|
var i = Math.floor((nn + 1) * pp),
|
|
y = B(i, nn - i + 1)();
|
|
if (y <= pp) {
|
|
acc += i;
|
|
nn -= i;
|
|
pp = (pp - y) / (1 - y);
|
|
} else {
|
|
nn = i - 1;
|
|
pp /= y;
|
|
}
|
|
}
|
|
var sign = pp < 0.5,
|
|
pFinal = sign ? pp : 1 - pp,
|
|
g = G(pFinal);
|
|
for (var s = g(), k = 0; s <= nn; ++k) s += g();
|
|
return acc + (sign ? k : nn - k);
|
|
};
|
|
}
|
|
|
|
randomBinomial.source = sourceRandomBinomial;
|
|
|
|
return randomBinomial;
|
|
})(defaultSource);
|
|
|
|
var weibull = (function sourceRandomWeibull(source) {
|
|
function randomWeibull(k, a, b) {
|
|
var outerFunc;
|
|
if ((k = +k) === 0) {
|
|
outerFunc = x => -Math.log(x);
|
|
} else {
|
|
k = 1 / k;
|
|
outerFunc = x => Math.pow(x, k);
|
|
}
|
|
a = a == null ? 0 : +a;
|
|
b = b == null ? 1 : +b;
|
|
return function() {
|
|
return a + b * outerFunc(-Math.log1p(-source()));
|
|
};
|
|
}
|
|
|
|
randomWeibull.source = sourceRandomWeibull;
|
|
|
|
return randomWeibull;
|
|
})(defaultSource);
|
|
|
|
var cauchy = (function sourceRandomCauchy(source) {
|
|
function randomCauchy(a, b) {
|
|
a = a == null ? 0 : +a;
|
|
b = b == null ? 1 : +b;
|
|
return function() {
|
|
return a + b * Math.tan(Math.PI * source());
|
|
};
|
|
}
|
|
|
|
randomCauchy.source = sourceRandomCauchy;
|
|
|
|
return randomCauchy;
|
|
})(defaultSource);
|
|
|
|
var logistic = (function sourceRandomLogistic(source) {
|
|
function randomLogistic(a, b) {
|
|
a = a == null ? 0 : +a;
|
|
b = b == null ? 1 : +b;
|
|
return function() {
|
|
var u = source();
|
|
return a + b * Math.log(u / (1 - u));
|
|
};
|
|
}
|
|
|
|
randomLogistic.source = sourceRandomLogistic;
|
|
|
|
return randomLogistic;
|
|
})(defaultSource);
|
|
|
|
var poisson = (function sourceRandomPoisson(source) {
|
|
var G = gamma.source(source),
|
|
B = binomial.source(source);
|
|
|
|
function randomPoisson(lambda) {
|
|
return function() {
|
|
var acc = 0, l = lambda;
|
|
while (l > 16) {
|
|
var n = Math.floor(0.875 * l),
|
|
t = G(n)();
|
|
if (t > l) return acc + B(n - 1, l / t)();
|
|
acc += n;
|
|
l -= t;
|
|
}
|
|
for (var s = -Math.log1p(-source()), k = 0; s <= l; ++k) s -= Math.log1p(-source());
|
|
return acc + k;
|
|
};
|
|
}
|
|
|
|
randomPoisson.source = sourceRandomPoisson;
|
|
|
|
return randomPoisson;
|
|
})(defaultSource);
|
|
|
|
// https://en.wikipedia.org/wiki/Linear_congruential_generator#Parameters_in_common_use
|
|
const mul = 0x19660D;
|
|
const inc = 0x3C6EF35F;
|
|
const eps = 1 / 0x100000000;
|
|
|
|
function lcg(seed = Math.random()) {
|
|
let state = (0 <= seed && seed < 1 ? seed / eps : Math.abs(seed)) | 0;
|
|
return () => (state = mul * state + inc | 0, eps * (state >>> 0));
|
|
}
|
|
|
|
exports.randomBates = bates;
|
|
exports.randomBernoulli = bernoulli;
|
|
exports.randomBeta = beta;
|
|
exports.randomBinomial = binomial;
|
|
exports.randomCauchy = cauchy;
|
|
exports.randomExponential = exponential;
|
|
exports.randomGamma = gamma;
|
|
exports.randomGeometric = geometric;
|
|
exports.randomInt = int;
|
|
exports.randomIrwinHall = irwinHall;
|
|
exports.randomLcg = lcg;
|
|
exports.randomLogNormal = logNormal;
|
|
exports.randomLogistic = logistic;
|
|
exports.randomNormal = normal;
|
|
exports.randomPareto = pareto;
|
|
exports.randomPoisson = poisson;
|
|
exports.randomUniform = uniform;
|
|
exports.randomWeibull = weibull;
|
|
|
|
Object.defineProperty(exports, '__esModule', { value: true });
|
|
|
|
})));
|