[Rd] fisher.g.test.single cannot handle time series longer than 341 (PR#7782)

btyner at stat.purdue.edu btyner at stat.purdue.edu
Fri Apr 8 03:42:58 CEST 2005


Full_Name: Benjamin Tyner
Version: R version 2.0.1, 2004-11-15
OS: Linux (kernel version 2.6.10-1.770_FC3)
Submission from: (NULL) (4.64.8.220)


On my build, for example:

library(GeneTS)
fisher.g.test.single(rnorm(341)) #this works fine
fisher.g.test.single(rnorm(342)) #this fails with the following error:

"Error in if (pval > 1) pval <- 1 : missing value where TRUE/FALSE needed"

The error always occurs for n>=342. This also affects the behavior of
fisher.g.test. I suspect the problem may lie in periodogram.spec



More information about the R-devel mailing list