/ndarray-sample-normalize

An adaption of scijs/ndarray-normalize using sample (n-1) statistics

Primary LanguageJavaScriptMIT LicenseMIT

ndarray-sample-normalize

Normalizes an ndarray so that the mean is 0 and the sample standard deviation is 1.

Example

var moments = require("ndarray-moments")
var baboon = require("luminance")(require("baboon-image"))
var baboon2 = require("luminance")(require("baboon-image"))

console.log("Before normalization:", moments(2, baboon))

console.log("\nSample stddev normalize")
require("../normalize")(baboon)

console.log("After sample normalization:", moments(2, baboon))

console.log("\nReqular normalize for comparison")
require("ndarray-normalize")(baboon2)

console.log("After regular normalization:", moments(2, baboon2))

This prints out:

Before normalization: [ 129.61546083073952, 18590.46474900778 ]

Sample stddev normalize
After sample normalization: [ -1.273097946508095e-10, 0.9999961854456416 ]

Reqular normalize for comparison
After regular normalization: [ 4.570595291131241e-11, 1.0000000001111533 ]

Install

npm install ndarray-sample-normalize

require("ndarray-sample-normalize")(output, input)

Normalizes an ndarray

  • output gets the result
  • input is the array that gets normalized (if unspecified, output is normalized in place)

Returns output

Credits

(c) 2013 Mikola Lysenko. MIT License adapted 2018 Baldur van Lew