Skip to content

Using permax with Data Frame Containing Missing Values

2 messages · Brian Lane, Robert Gentleman

#
I'm new to this site so I hope this isn't too naive a problem. I'm trying 
to use the permax function with a data frame containing gene expression 
measurements taken from 79 microarray experiments with >3000 genes per 
array. The data contains missing values and every time I use permax with 
the data frame I get the error:

NA/NaN/Inf in foreign function call (arg 1)

Could anyone suggest how I might get round this problem?

Regards,
Brian Lane
Dept of Haematology
University of Liverpool
#
There is currently no handling of NAs in permax. Your only simple
option is to drop those rows with NA's in them, or to perform some
sort of imputation.

I will mention it to the package's author, 
 Robert
On Wed, Jul 07, 2004 at 12:47:05PM +0100, Brian Lane wrote: