box_cox_transform: Box-Cox Transformation

In kota7/MLPipe: Machine Learning Pipeline

Description

Conduct Box-Cox Transformation. This can be used as

• an estimation process where data are fitted to the linear regression model of the transformed variable; or

• a transformer that transforms dependent variable

Transformation parameters can be fixed by users, or estimated by the maximum likelihood.

Usage

box_cox_transform(lambda = 1, lambda2 = NULL,
  skipfit = FALSE, tol = 1e-5)

Arguments

lambda

initial value for lambda parameter

lambda2

initial value for lambda2 parameter or logical that indicates if lambda2 should be estimated. If NULL or FALSE, then lambda2 is fixed to 0

skipfit

logical. If TRUE, fit method does nothing and the parameters are fixed to the initial values.

tol

lambda smaller than this level is regarded as 0 and log function is applied

Value

BoxCoxTransform class object

Class Attributes

beta

regression coefficients

Class Methods

fit(x = NULL, y)

if skipfit is FALSE, then estimate the lambda parameter(s) by the maximum likelihood, otherwise, the parameters are fixed. In either case, regression coeffients beta are estimated by the least squares

transform(x = NULL, y)

transform y and returns list of x and y

inv_transform(x = NULL, y)

inverse transform y and returns list of x and y

predict(x, ...)

return predicted values of y in the pre-transfom scale

incr_fit(x, y)

not available

Details

uses boxcoxfit as the backend paramter estimator

Examples

simul <- geoR::rboxcox(100, lambda=0.5, mean=10, sd=2)
b <- box_cox_transform()
b$fit(y=simul)
cat(b$lambda, '\n')

b <- box_cox_transform(lambda2=TRUE)
b$fit(y=seq(-1, 1, 1/20))
cat(b$lambda, b$lambda2, '\n')

data(trees)
b <- box_cox_transform()
x <- trees[,1:2]
y <- trees[,3]
b$fit(x, y)
pred <- b$predict(x)$y
cor(y, pred)
## Not run: 
plot(y, pred)
## End(Not run)

b <- box_cox_transform(lambda=0, skipfit=TRUE)
b$fit(y=1:10)
cat(b$lambda, b$beta, '\n')

kota7/MLPipe documentation built on May 5, 2019, 5:53 p.m.