statty.js

Statistics for javascript

Intended for node.js. Available through npm. ###Installation npm install statty.js

Or clone the repo and put it in your project.

###Setup

var stats = require('statty.js')
console.log(stats.normal(5,1).rand())

So far the only distributions are the normal, uniform, laplace, poisson, pareto, exponential, geometric, bernoulli, and binomial. In general, each distribution is initalized with the parameters listed on Wikipedia, in the order listed there. When using the fit method, which is available for all distributions expcept the binomial, parameters are calculated using the Maximum Likelihood Estimator for the distribution.

#####Examples

var stats = require('statty.js')
norm = stats.normal(5,1)            \\ normal with mean 5, variance 1
unif = stats.uniform(10,20)         \\ uniform from range 10 to 20
pare = stats.pareto(1,6/5)          \\ pareto with scale 1 and shape 1.2 
lapl = stats.laplace(10,4)          \\ laplace with mean 10, scale 4 
geom = stats.geometric(.5)          \\ geometric with sucess .5
pois = stats.poisson(10)            \\ poisson with mean 10 
expo = stats.exponential(1/10)      \\ exponential with mean 10 
bern = stats.bernoulli(.7)          \\ bernoulli with mean .7
bino = stats.binomial(10,.7)        \\ binomial with 10 trials, probability .7
console.log(norm.pdf(1))            \\ probability density function
console.log(bern.pmf(1))            \\ discrete distributions have pmf
console.log(norm.cdf(4))            \\ cumulative density function
console.log(norm.quantile(.9))      \\ quantile
console.log(bern.rand())            \\ generates a random bernoulli trial
\\or
norm = stats.normal.fit([1,2,3,4])  \\ returns model fitted to data
console.log(norm.mean)              \\ mean attribute (calculated for uniform)
console.log(norm.variance)          \\ variance attribute (calculated for uniform)
console.log(norm.rand(10))          \\ generates array of 10 random numbers