R: Transform variables based on the output of 'transfo'.

transfo_newdata

R Documentation

Transform variables based on the output of `transfo`.

Description

Based on the output of transfo, transform the variables using Yeo-Johnson and/or Box-Cox transformations with the previously estimated parameters and standardization.

Usage

transfo_newdata(Xnew, transfo.out)

Arguments

`Xnew`	A data matrix with d columns, which contain the variables to be transformed. The number of columns and their names must be the same as those of the original data on which `transfo` was run. The number of rows may be different.
`transfo.out`	The output of a call to `transfo`.

Value

Returns a matrix with transformed variables.

Author(s)

J. Raymaekers and P.J. Rousseeuw

References

J. Raymaekers and P.J. Rousseeuw (2021). Transforming variables to central normality. Machine Learning. \Sexpr[results=rd,stage=build]{tools:::Rd_expr_doi("10.1007/s10994-021-05960-5")}(link to open access pdf)

Examples

set.seed(123); tempraw <- matrix(rnorm(2000), ncol = 2)
tempx <- cbind(tempraw[, 1],exp(tempraw[, 2]))
tempy <- 0.5 * tempraw[, 1] + 0.5 * tempraw[, 2] + 1
x <- tempx[1:900, ]
y <- tempy[1:900]
tx.out <- transfo(x, type = "bestObj")
tx.out$ttypes
tx.out$lambdahats
tx <- tx.out$Y
lm.out <- lm(y ~ tx)
summary(lm.out)
xnew <- tempx[901:1000, ]
xtnew <- transfo_newdata(xnew, tx.out)
yhatnew <- tcrossprod(lm.out$coefficients, cbind(1, xtnew)) 
plot(tempy[901:1000], yhatnew); abline(0, 1)

Transform variables based on the output of transfo.