recommend.box.cox: Normalizing Transformations - Box-Cox/Ladder of Powers -...

In burrm/lolcat: Miscellaneous Useful Functions

Normalizing Transformations - Box-Cox/Ladder of Powers - Parameter Recommendation

Description

Recommend parameters to optimally transform sample data to normality based on the Box Cox transformation.

Usage

recommend.box.cox(
  x,
  lambda.min = -5,
  lambda.max = 5,
  step = 0.1,
  correct.min = T,
  target = c("normal"),
  metric = c("skewness+kurtosis", "skewness", "kurtosis")
)

Arguments

x	Vector/numeric - sample data.
lambda.min	Scalar/numeric - Minimum exponent.
lambda.max	Scalar/numeric - Maximum exponent.
step	- Scalar/numeric - The step size between min and max exponent to try.
correct.min	Scalar/logical - Shift data by minimum in sample to avoid transformation/reverse transformation issues.
target	- Scalar/character - The target distribution.
metric	- Scalar/character - Which metric to use for evaluating transformations.

Value

Data frame with best parameters found in search.

burrm/lolcat documentation built on Sept. 15, 2023, 11:35 a.m.