[BioC] liandwong and mas5

[Hee Siew Wan]

I've tried with larger probe pairs (207149 of ath1121501 array) but have the same problem. However, there are some probe sets with only 4 pairs. Could it be because of this small number of pairs per set?
 
When I use traceback();
> traceback()
12: stop("need at least four unique `x' values")
11: smooth.spline(ref[i.set], data[i.set])
 
Does it mean that I need at least 4 intensity values?

	-----Original Message----- 
	From: Laurent Gautier [mailto:lgautier at altern.org] 
	Sent: Wed 17-Nov-04 10:38 PM 
	To: Hee Siew Wan 
	Cc: bioconductor at stat.math.ethz.ch 
	Subject: Re: [BioC] liandwong and mas5
	
	

	Hee Siew Wan wrote:
	> Dear All,
	> 
	> I created a new cdf for the Dilution data:
	> data(Dilution)
	> i1 <- 13:16
	> i2 <- 29:32
	> i3 <- 45:48
	> i4 <- 61:64
	> i5 <- 77:80
	> i6 <- 93:96
	> m <- list(match = list(i1, i2, i3, i4, i5, i6))
	> ids <- unique(hgu95av2probe$Probe.Set.Name[1:96])
	> alt.cdf <- buildCdfEnv.matchprobes(m, ids, nrow.chip=640,
	> ncol.chip=640, chiptype="HG-U95Av2", probes.pack="hgu95av2probe")
	> new.dil <- Dilution[,1:2]
	> validAffyBatch(new.dil, alt.cdf)
	> new.dil.cdfenv <- alt.cdf at envir
	> new.dil at cdfName <- "new.dil.cdfenv"
	> 
	> I encountered some problems when I tried to use the following expresso:
	> 
	>
	>>dil.lw <- liandwong(new.dil)
	>
	> normalization: invariantset
	> PM/MM correction : subtractmm
	> expression values: liwong
	> normalizing...Error in smooth.spline(ref[i.set], data[i.set]) :
	>         need at least four unique `x' values
	> 
	>
	>>dil.mas <- mas5(new.dil)
	>
	> background correction: mas
	> PM/MM correction : mas
	> expression values: mas
	> background correcting...Error in as.vector(data) : NA/NaN/Inf in foreign function call (arg 1)
	> 
	> What do those errors mean?
	
	The errors are emitted by the functions doing the job behind the
	wrappers... a call to 'traceback' call tell you more...
	
	I would expect the pre-processing method you use to fail (or give
	meaningless results) with such a small number of probes in use: you are
	only handling 24 probes in your CDF !
	
	
	> I'm using:
	> R version 2.0.0
	> affy version 1.5.8
	> altcdfenvs version 1.1-9
	> hgu95av2probe version 1.0
	> 
	> I'd appreciate any help. Thank you.
	> 
	> Regards
	> SiewWan
	>
	> _______________________________________________
	> Bioconductor mailing list
	> Bioconductor at stat.math.ethz.ch
	> https://stat.ethz.ch/mailman/listinfo/bioconductor
	>