Function

Static Public Summary
public	avgPoints(pointArr: array, pointSums: array): array Returns the average of x values and y values in a 2 dimensional array
public	balanceMagnitude(result: number, variable: number, weight: number, error: number): number Returns a value that, when multiplied with variable, weight, and error arguments, will result in a value of the same magnitude as the result argument.
public	balanceMagnitudeArray(inputArr: array[number], weightArr: array[number], error: number): array[number] Returns an array of values that, when multiplied with variable, weight, and error arguments, will result in a value of the same magnitude as the result argument.
public	cummulativeAvgPoints(input: number \| array[numbers], cummulativeAvg: number, cummulativeN: number): number Returns average value of the given input and a previously calculated average
public	cummulativeRangePoints(input: number \| array[numbers], cummulativeRange: array[numbers]): array Returns range of the given inputs and a previously calculated range
public	evalError(inputArr: array[numbers], weightArr: array[numbers]): number Returns an error value from inputs and weights
public	evalGuess(input: array[numbers], weightArr: array[numbers]): number Returns an expected result using supplied inputs and weights
public	gradientStep(weightArr: array[numbers], input: array[numbers], error: number, learningRateArr: array[number]): array Returns a new set of weights
public	interceptOLS(lineObj: object, avg: array): number Returns the intercept for a line of an Ordinary Least Squares regression
public	meanSquaredErrors(pointArr: array, lineObj: object): number Returns the MSE of a 2D coordinate array
public	ols_regression(pointArr: *): object Returns an object representing the ordinary least squares line
public	sigmoidCurve(x: number): number Returns the position x in a sigmoid curve
public	sigmoidDerivative(x: number): number Returns the position x in a derivative sigmoid curve
public	sumPoints(pointArr: array): array Returns the sum of x values and y values in a 2 dimensional array
public	sumSquaredErrors(pointArr: array, lineObj: object): number Returns the SSE (sum of squared errors) for a 2D coordinate array
public	sumSquaredResiduals(pointArr: array, avg: array): number Returns the SSR (sum of squared residuals) for a 2D coordinate array

Static Public

public avgPoints(pointArr: array, pointSums: array): array source

import {avgPoints} from 'viktor/lib/utils/1D/avgPoints/avgPoints.js'

Returns the average of x values and y values in a 2 dimensional array

Params:

Name	Type	Attribute	Description
pointArr	array		An array containing all coordinate points
pointSums	array		An array containing the computed sum of all x values and all y values

Return:

array

An array containing the average x value and y value of the pointArr

public balanceMagnitude(result: number, variable: number, weight: number, error: number): number source

import {balanceMagnitude} from 'viktor/lib/utils/0D/balanceMagnitude/balanceMagnitude.js'

Returns a value that, when multiplied with variable, weight, and error arguments, will result in a value of the same magnitude as the result argument.

Params:

Name	Type	Attribute	Description
result	number		a y value, or the variable that the line is fitted to predict.
variable	number		an x value.
weight	number		the weight associated with the x value.
error	number		the error associated with the current gradient step.

Return:

number

An integer representing the power of 10 value that would balance result with the product of all other arguments

public balanceMagnitudeArray(inputArr: array[number], weightArr: array[number], error: number): array[number] source

import {balanceMagnitudeArray} from 'viktor/lib/utils/1D/balanceMagnitudeArray/balanceMagnitudeArray.js'

Returns an array of values that, when multiplied with variable, weight, and error arguments, will result in a value of the same magnitude as the result argument.

Params:

Name	Type	Attribute	Description
inputArr	array[number]		an array of y and x values representing a potential input.
weightArr	array[number]		an array of weights.
error	number		the error associated with the current gradient step.

Return:

array[number]

An array representing the power of 10 values that would balance results with the product of all other arguments

public cummulativeAvgPoints(input: number | array[numbers], cummulativeAvg: number, cummulativeN: number): number source

import {cummulativeAvgPoints} from 'viktor/lib/utils/1D/cummulativeAvgPoints/cummulativeAvgPoints.js'

Returns average value of the given input and a previously calculated average

Params:

Name	Type	Attribute	Description
input	number \| array[numbers]		A number or array representing the values to be averaged
cummulativeAvg	number		A value representing the current average of values seen up until this point
cummulativeN	number		A value representing the number of values seen up until this point

Return:

number

A weighted average

public cummulativeRangePoints(input: number | array[numbers], cummulativeRange: array[numbers]): array source

import {cummulativeRangePoints} from 'viktor/lib/utils/1D/cummulativeRangePoints/cummulativeRangePoints.js'

Returns range of the given inputs and a previously calculated range

Params:

Name	Type	Attribute	Description
input	number \| array[numbers]		A number or array representing the values to be evaluated
cummulativeRange	array[numbers]		A array representing the current range of values seen up until this point

Return:

array

A range of the given inputs

public evalError(inputArr: array[numbers], weightArr: array[numbers]): number source

import {evalError} from 'viktor/lib/utils/1D/evalError/evalError.js'

Returns an error value from inputs and weights

Params:

Name	Type	Attribute	Description
inputArr	array[numbers]		An array representing the values to be evaluated, with the y value in the [0] position
weightArr	array[numbers]		An array representing the weights to be used with the values, with b0 in the [0] position

Return:

number

Sum of errors using the weightArr

public evalGuess(input: array[numbers], weightArr: array[numbers]): number source

import {evalGuess} from 'viktor/lib/utils/1D/evalGuess/evalGuess.js'

Returns an expected result using supplied inputs and weights

Params:

Name	Type	Attribute	Description
input	array[numbers]		An array representing x and y values
weightArr	array[numbers]		An array representing weights used to estimate a y value

Return:

number

An expected y value

public gradientStep(weightArr: array[numbers], input: array[numbers], error: number, learningRateArr: array[number]): array source

import {gradientStep} from 'viktor/lib/utils/1D/gradientStep/gradientStep.js'

Returns a new set of weights

Params:

Name	Type	Attribute	Description
weightArr	array[numbers]		An array of numbers representing estimated weights, with b0 in the [0] position
input	array[numbers]		An array of numbers representing the original inputs in some way, with y in the [0] position
error	number		A number representing an error value pointing in the direction of the next gradient step
learningRateArr	array[number]		An array of numbers representing the magnitudes of changes in the weights

Return:

array

A new set of weights

public interceptOLS(lineObj: object, avg: array): number source

import {interceptOLS} from 'viktor/lib/ordinaryLeastSquares/interceptOLS/interceptOLS.js'

Returns the intercept for a line of an Ordinary Least Squares regression

Params:

Name	Type	Attribute	Description
lineObj	object		An object containing at least a b1 property with the value of the expected slope
avg	array		An array containing a y value at [1] and an x value at position [0]

Return:

number

An expected intercept value

public meanSquaredErrors(pointArr: array, lineObj: object): number source

import {meanSquaredErrors} from 'viktor/lib/utils/1D/meanSquaredErrors/meanSquaredErrors.js'

Returns the MSE of a 2D coordinate array

Params:

Name	Type	Attribute	Description
pointArr	array		An array containing all coordinate points
lineObj	object		An object containing the b1 and b0 weights for an OLS regression

Return:

number

A number representing the MSE of a 2D coordinate array

public ols_regression(pointArr: *): object source

import {ols_regression} from 'viktor/lib/ordinaryLeastSquares/ols_regression/ols_regression.js'

Returns an object representing the ordinary least squares line

Params:

Name	Type	Attribute	Description
pointArr	*

Return:

object

An object representing a line

public sigmoidCurve(x: number): number source

import {sigmoidCurve} from 'viktor/lib/curves/sigmoidCurve/sigmoidCurve.js'

Returns the position x in a sigmoid curve

Params:

Name	Type	Attribute	Description
x	number		An x value along the sigmoid curve

Return:

number

y - The y value corresponding to the provided x value

public sigmoidDerivative(x: number): number source

import {sigmoidDerivative} from 'viktor/lib/curves/sigmoidDerivative/sigmoidDerivative.js'

Returns the position x in a derivative sigmoid curve

Params:

Name	Type	Attribute	Description
x	number		An x value along the derivative sigmoid curve

Return:

number

y - The y value corresponding to the provided x value

public sumPoints(pointArr: array): array source

import {sumPoints} from 'viktor/lib/utils/1D/sumPoints/sumPoints.js'

Returns the sum of x values and y values in a 2 dimensional array

Params:

Name	Type	Attribute	Description
pointArr	array		An array containing all coordinate points

Return:

array

An array containing the sums of x values and y values of the pointArr

public sumSquaredErrors(pointArr: array, lineObj: object): number source

import {sumSquaredErrors} from 'viktor/lib/utils/1D/sumSquaredErrors/sumSquaredErrors.js'

Returns the SSE (sum of squared errors) for a 2D coordinate array

Params:

Name	Type	Attribute	Description
pointArr	array		An array containing a set of coordinate points
lineObj	object		An object containing the b1 and b0 weights for an OLS regression

Return:

number

A number representing the SSE of the points in pointArr given the line specified by lineObj

public sumSquaredResiduals(pointArr: array, avg: array): number source

import {sumSquaredResiduals} from 'viktor/lib/utils/1D/sumSquaredResiduals/sumSquaredResiduals.js'

Returns the SSR (sum of squared residuals) for a 2D coordinate array

Params:

Name	Type	Attribute	Description
pointArr	array		An array containing a set of coordinate points
avg	array		An array containing the average of all y values and all x values

Return:

number

A number representing the SSR of the points in pointArr given the average value