COINr (version 1.1.14)

qNormalise.data.frame: Quick normalisation of a data frame

Description

This is a wrapper function for Normalise(), which offers a simpler syntax but less flexibility. It normalises a data frame using a specified function f_n which is used to normalise each column, with additional function arguments passed by f_n_para. By default, f_n = "n_minmax" and f_n_para is set so that the columns of x are normalised using the min-max method, between 0 and 100.

Usage

# S3 method for data.frame
qNormalise(x, f_n = "n_minmax", f_n_para = NULL, directions = NULL, ...)

Value

A normalised data frame

Arguments

x

A numeric data frame

f_n

Name of a normalisation function (as a string) to apply to each column of x. Default "n_minmax".

f_n_para

Any further arguments to pass to f_n, as a named list. If f_n = "n_minmax", this defaults to list(l_u = c(0,100)) (scale between 0 and 100).

directions

An optional data frame containing the following columns:

• iCode The indicator code, corresponding to the column names of the data frame

• Direction numeric vector with entries either -1 or 1 If directions is not specified, the directions will all be assigned as 1. Non-numeric columns do not need to have directions assigned.

...

arguments passed to or from other methods.

Details

Essentially, this function is similar to Normalise() but brings parameters into the function arguments rather than being wrapped in a list. It also does not allow individual normalisation.

See Normalise() documentation for more details, and vignette("normalise").

Examples

# some made up data
X <- data.frame(uCode = letters[1:10],
                a = runif(10),
                b = runif(10)*100)
# normalise (defaults to min-max)
qNormalise(X)