[BioC] Handling Missing Values

Stephen Henderson s.henderson@ucl.ac.uk
Wed, 30 Oct 2002 09:24:13 -0000

In the PAML package for R of Tibshirani (Shrunken Centroid Learning) they
have used K-nearest neighbour imputation. Obviously you should not use this
if there are too many missing values, but then most sensible people will
choose to reject an experiment if it is too corrupted prior to this.

Stephen Henderson
Wolfson institute for Biomedical Research
UCL, Gower Street, WC1E 6BT, UK
44 (0)207 679 6827

-----Original Message-----
From: Robert Gentleman
To: Sundaram, Shyam (NIH/CIT)
Cc: 'bioconductor@stat.math.ethz.ch'
Sent: 29/10/02 20:56
Subject: Re: [BioC] Handling Missing Values

On Tue, Oct 29, 2002 at 12:33:17PM -0500, Sundaram, Shyam (NIH/CIT)
> Hi,
> Couple of  questions regarding the bioconductor packages and the
handling of
> missing values.
> 1. Is there a currently an impute function implementation( or
likelihood of
> adding an impute function)?

  It is unclear what such a function would do. In most cases one is
  going to have to use some reasonable amount of external information
  to perform imputations. This is possible now.

> 2. Does the exprSet object have a restriction of having missing values
> the "exprs" slot ?

  Not intentionally (and not that I am aware of). If you have a
  specific example please post it. If you are simply wondering then I
  suggest that you actually try it and find out. There are lots of
  runnable examples that can easily be adjusted to produce missing
  values and without too much effort you can see the answer for


> Thanks
> Shyam
> _______________________________________________
> Bioconductor mailing list
> Bioconductor@stat.math.ethz.ch
> http://www.stat.math.ethz.ch/mailman/listinfo/bioconductor

| Robert Gentleman                 phone : (617) 632-5250
| Associate Professor              fax:   (617)  632-2444
| Department of Biostatistics      office: M1B20
| Harvard School of Public Health  email:
rgentlem@jimmy.dfci.harvard.edu   |
Bioconductor mailing list

This email and any files transmitted with it are confidential and
intended solely for the use of the individual or entity to whom they
are addressed. If you have received this email in error please notify
the system manager (wibr.mail@ucl.ac.uk). All files are scanned for viruses.