martedì 29 maggio 2007

Detecting outliers through boxplots of the features

This function detects univariate outliers simultaneously using boxplots
of the features:

require(dprep)

data(diabetes)

outbox(diabetes,nclass=1)

7 commenti:

  1. Where can we find the 'dprep' package?

    RispondiElimina
  2. Unfortunately it seems that the dprep package was removed from the CRAN repository. Previous versions of the package can be downloaded from the following link:
    http://cran.stat.ucla.edu/web/packages/dprep/index.html

    RispondiElimina
  3. Why has the package been removed from the CRAN repository?

    RispondiElimina
  4. I guess the authors can maintain the software no longer.

    RispondiElimina
  5. Is there any alternative for this package, which does the similar work??

    RispondiElimina
  6. Not sure it is a real alternative, but the Bioconductor package parody (http://www.bioconductor.org/packages/2.11/bioc/html/parody.html) presents routines for univariate and multivariate outlier detection that can be of some use.
    HIH

    RispondiElimina