bigO.drop: Ben Lira's function for assessing and dropping multivariate...

Description Usage Arguments Value

View source: R/tidying.r

Description

The function provides information on multivariate outliers, but also produces a dataset with multivariate outliers excluded, which can be saved to a new object, or used to overwrite the data set. This calculates Mahalanobis distance for each observation and calculates multivariate outliers using Chi2 testing.

Usage

1
bigO.drop(data, vars = NULL, exclude = NULL, p = 0.001)

Arguments

data

data frame

vars

vector of subset of data to analyze. Takes either varnames or a numeric vector of column numbers.

exclude

vector of subset of data to exclude from analysis. Takes either varnames or a numeric vector of column numbers.

p

level at which to test the significance of outliers, by testing the Chi2 parameter of the alpha distribution.

Value

cases excluded, a dataset of remaining cases, and a graph showing excluded vs. retained cases.


crbwin/clnR documentation built on Oct. 29, 2020, 1:55 a.m.