mark (version 0.8.2)

normalize: Normalize values

Description

Normalizes values based on possible range and new bounds

Usage

normalize(x, ...)
# S3 method for default
normalize(x, range = base::range(x, na.rm = TRUE), bounds = 0:1, ...)
# S3 method for data.frame
normalize(x, ...)

Value

x with transformed values where range values are transformed to bounds.

Arguments

x

An object that is (coercible to) double; data.frames are transformed

...

Additional arguments passed to methods

range

The range of possible values of x. See details for more info. Defaults to the range of non-NA values

bounds

The new boundaries for the normalized values of x. Defaults to 0 and 1.

Details

Parameters range and bounds are modified with base::range(). The largest and smallest values are then used to determine the minimum/maximum values and lower/upper bounds. This allows for a vector of more than two values to be passed.

The current implementation of normalize.data.frame() allows for list of parameters passed for each column. However, it is probably best suited for default values.

Examples

x <- c(0.23, 0.32, 0.12, 0.61, 0.26, 0.24, 0.23, 0.32, 0.29, 0.27)
data.frame(
  x  = normalize(x),
  v  = normalize(x, range = 0:2),
  b  = normalize(x, bounds = 0:10),
  vb = normalize(x, range = 0:2, bounds = 0:10)
)

# maintains matrix
mat <- structure(c(0.24, 0.92, 0.05, 0.37, 0.19, 0.69, 0.43, 0.22, 0.85,
0.73, 0.89, 0.68, 0.57, 0.89, 0.61, 0.98, 0.75, 0.37, 0.24, 0.24,
0.34, 0.8, 0.25, 0.46, 0.03, 0.71, 0.79, 0.56, 0.83, 0.97), dim = c(10L, 3L))

mat
normalize(mat, bounds = -1:1)
normalize(as.data.frame(mat), bounds = -1:1)