efuns

README

Docs: GoDoc

Package efuns has misc functions, such as Gaussian and Logistic, that are used in neural models, and do not have a home elsewhere.

Documentation

Overview

Package efuns has misc functions, such as Gaussian and Logistic, that are used in neural models, and do not have a home elsewhere.

Index

• func Gauss1DNoNorm(x, sig float32) float32
• func GaussVecDistNoNorm(a, b math32.Vector2, sigma float32) float32
• func Logistic(x, gain, off float32) float32

Functions

func Gauss1DNoNorm

func Gauss1DNoNorm(x, sig float32) float32

Gauss1DNoNorm returns the gaussian of a given x value, without normalizing (i.e., max value is 1 at x = 0)

func GaussVecDistNoNorm

func GaussVecDistNoNorm(a, b math32.Vector2, sigma float32) float32

GaussVecDistNoNorm returns the gaussian of the distance between two 2D vectors using given sigma standard deviation, without normalizing area under gaussian (i.e., max value is 1 at dist = 0)

func Logistic

func Logistic(x, gain, off float32) float32

Logistic is the logistic (sigmoid) function of x: 1/(1 + e^(-gain*(x-off)))