| standardize {exploreR} | R Documentation |
Standardize Variables
Description
This function takes in a dataframe, the name of any number of variables. It then returns either a vector or a dataframe of scaling results. If passed a single variable name, standardize will return a the standardized variable as a vector, when passed 2 or more variable names, standardize will return a data frame containing all of the standardized variables.
Usage
standardize(data, variable, type = "absolute")
Arguments
data |
data.frame object that contains both the dependent variable and predictor variables you want to regress. |
variable |
variable name or vector of names for variables you want standardized. |
type |
by default, 'absolute' will scale every variable from 0 to 1. "classic" is a little more complicated where the variable is rescaled the mean equaling 0 and a standard deviation is 1. |
Details
Often times we are forced to compare variables which exist on different scales, but how do you compare the coefficient for a country's population to one that's much smaller? Standardizing your variables can make reading regression results more useful because it will make coeficients more comparable.
Value
if the function is passed a single variable to standardize, it will return a vector of all obeservations in the variable standardized. If the function is passed more than one variable name, it will return a data-frame containing the new observation values.
Examples
exam.df <- iris
standardize(exam.df, "Petal.Width")
standardize(exam.df, c("Petal.Width", "Petal.Length"), type = "classic")