[BioC] strange results with edgeR::goodTuring

[Francois Pepin]

Hi everyone,

I'm trying to use the goodTuring function in edgeR to estimate what kind of pseudocounts to use and I'm getting strange results with small number of categories:

x<-c(312,14491,16401,65124,129797,323321,366051,368599,405261,604962)
y<- goodTuring(x)
y
$count
 [1]    312  14491  16401  65124 129797 323321 366051 368599 405261 604962

$proportion
 [1] 0 0 0 0 0 0 0 0 0 1

$P0
[1] 0

$n0
[1] 0


If I'm understanding this properly, y$proportion is telling me that I should expect all my counts to fall under the last category, which does not make sense. I would expect something pretty close to x/sum(x) instead.

This is a bit of a toy example and I'm mostly interested in cases where I have more categories but it would be nice if this could work in all cases.

sessionInfo()
R version 2.15.1 (2012-06-22)
Platform: x86_64-unknown-linux-gnu (64-bit)

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=C                 LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] edgeR_2.6.9   limma_3.12.1  dataframe_2.5


Thanks,

François