mctad.js

About

mctad.js is a JavaScript library for probability and statistics. It was created to provide functionality missing from other such libraries — such as having an awareness of many common discrete and continuous probability distributions, and the ability to generate random variables from those distributions — although it also provides a growing list of standard descriptive statistics and statistical analyses.

Descriptive Statistics

Given an Array of data, mctad.js can calculate its

• arithmetic_mean(Array), aliased as mean(Array)
• geometric_mean(Array)
• median(Array)
• mode(Array), always returned as an Array, if defined
• sample_variance(Array)
• sample_standard_deviation(Array)

There are also public methods for sum and product.

> data = [1, 3, 6, 6, 6, 6, 7, 7, 12, 12, 12, 12, 17];
  [1, 3, 6, 6, 6, 6, 7, 7, 12, 12, 12, 12, 17]
> mctad.sum(data);
  107
> mctad.mean(data);
  8.23076923076923
> mctad.median(data);
  7
> mctad.mode(data);
  [6, 12]

When defined and non-trivial to implement, descriptive statistics are returned with any probability distribution objects you create.

Confidence Intervals

mctad.js can create one- and two-tailed confidence intervals on the:

• population mean
• proportion
• difference between two means
• difference between two proportions

Probability Distributions

mctad.js can create many common discrete and continuous probability distributions from parameters you supply. The resulting probability distribution objects contain common statistics on the distribution (when defined); convenience methods for accessing probability distribution, probability mass, and cumulative probability functions; domain & range helpers to simplify integration with D3, and random variable generators.

• Discrete Probability Distributions
  • Discrete Uniform
  • Bernoulli
  • Binomial
  • Geometric
  • Hypergeometric
  • Pascal
  • Poisson
• Continuous Probability Distributions
  • Uniform
  • Triangular
  • Exponential
  • Normal
  • Lognormal
  • Student’s t

Core and Helpers

mctad.js has a number of methods, included in /src/core. mctad.js uses them internally, but they are available for your use, and include

• isInteger()
• allPositive(Array)
• sortNumeric(Array)
• toRadians(v)
• getBaseLog(x, y)
• getRandomArbitrary(min, max)
• getRandomInt(min, max)
• factorial(n)
• doubleFactorial(n)
• combination(n, k)
• Γ(n), aliased as gamma(n)

Documentation for these is interspersed with the code, meaning the Groc-generated documentation is the best place to see how they’re implemented and to learn what they do.

Documentation

Groc-generated documentation is available here and in the /doc directory of your cloned repository.