winsorize: Winsorize vector

In egenn/rtemis: Machine Learning and Visualization

Winsorize vector

Description

Replace extreme values by absolute or quantile threshold

Usage

winsorize(
  x,
  lo = NULL,
  hi = NULL,
  prob_lo = 0.025,
  prob_hi = 0.975,
  quantile_type = 7,
  verbosity = 1L
)

Arguments

x	Numeric vector: Input data
lo	Numeric: If not NULL, replace any values in x lower than this with this.
hi	Numeric: If not NULL, replace any values in x higher than this with this.
prob_lo	Numeric (0, 1): If not NULL and lo = NULL, find sample quantile that corresponds to this probability and set as lo.
prob_hi	Numeric (0, 1): If not NULL and hi = NULL, find sample quantile that corresponds to this probability and set as hi.
quantile_type	Integer: passed to stats::quantile
verbosity	Integer: Verbosity level.

Details

If both lo and prob_lo or both hi and prob_hi are NULL, cut-off is set to min(x) and max(x) respectively, i.e. no values are changed

Value

Numeric vector.

Author(s)

EDG

Examples

## Not run: 
# Winsorize a normally distributed variable
x <- rnorm(500)
xw <- winsorize(x)
# Winsorize an exponentially distributed variable only on
# the top 5% highest values
x <- rexp(500)
xw <- winsorize(x, prob_lo = NULL, prob_hi = .95)

## End(Not run)

egenn/rtemis documentation built on June 14, 2025, 11:54 p.m.