dropvar function

Drop variable

Drops columns in a matrix to avoid perfect multicollinearity.

dropvar(x, tol=1e-07, LAPACK=FALSE, silent=FALSE)

Arguments

• x: a matrix, possibly less than full column rank.
• tol: numeric value. The tolerance for detecting linear dependencies among regressors, see qr function. Only used if LAPACK is FALSE
• LAPACK: logical, TRUE or FALSE (default). If true use LAPACK otherwise use LINPACK, see qr function
• silent: logical, TRUE (default) or FALSE. Whether to print a notification whenever a regressor is removed

Returns

a matrix whose regressors linearly independent

Details

Original function drop.coef developed by Rune Haubo B. Christensen in package ordinal, https://cran.r-project.org/package=ordinal.

References

Rune H.B. Christensen (2014): 'ordinal: Regression Models for Ordinal Data'. https://cran.r-project.org/package=ordinal

Author(s)

Rune Haubo B. Christensen, with modifications by Genaro Sucarrat, http://www.sucarrat.net/

See Also

isat

Examples

set.seed(1)
x <- matrix(rnorm(20), 5)
dropvar(x) #full rank, none are dropped

x[,4] <- x[,1]*2
dropvar(x) #less than full rank, last column dropped