Class PCR

Creates new PCR (Principal component regression)

param predictor: matrix with predictor variables. (Each column is an array)
param response: matrix with response variables. (Each column is an array)
param pcaWeight: Weight to choose the principal components. It refers to the weight that components must sum with each other (in percent) to perform the regression.
param intercept: Intercept

Hierarchy

PCR

Index

Constructors

constructor

Properties

Methods

Constructors

constructor

new PCR(predictor: number[][] | Matrix, response: number[][] | Matrix, options?: PCROptions): PCR

- Defined in pcr.ts:28
Parameters
- predictor: number[][] | Matrix
- response: number[][] | Matrix
- options: PCROptions = {}
Returns PCR

Properties

coefficients

coefficients: Matrix

intercept

intercept: boolean

loadingsData

loadingsData: { component: number[]; componentNumber: number; evalues: number; weight: number }[]

pcaWeight

pcaWeight: number

scores

scores: Matrix

statistics

statistics: Statistics

Optional xMedia

xMedia?: number[]

yFittedValues

yFittedValues: Matrix

Methods

getCoefficients

getCoefficients(): Matrix

- Defined in pcr.ts:232
Returns the regression coefficients

Returns Matrix

getFittedValuesY

getFittedValuesY(): Matrix

- Defined in pcr.ts:224
Returns fitted values of Y

Returns Matrix

getLoadingsdata

getLoadingsdata(): { component: number[]; componentNumber: number; evalues: number; weight: number }[]

- Defined in pcr.ts:240
Returns the scores for principal components

Returns { component: number[]; componentNumber: number; evalues: number; weight: number }[]

getScores

getScores(): Matrix

- Defined in pcr.ts:248
Returns the number of principal components used

Returns Matrix

getStatistic

getStatistic(): Statistics

- Defined in pcr.ts:216
Returns some basic statistics of the regression

Returns Statistics

predict

predict(x: number[]): number[]

- Defined in pcr.ts:196
Predict y-values for a given x

Parameters
- x: number[]
Returns number[]