collapse-methods: collapse
In drmjc/microarrays: A collection of microarray and genomic helper code.

Description Arguments Details Value collapse,LumiBatch and column collapse,LumiBatch and Missing values Author(s) Examples

collapse a dataset to have just 1 row per unique value of key.

`x`	an object
`decreasing`	logical: should the sort order be increasing or decreasing?
`na.last`	for controlling the treatment of `NA`s. If `TRUE`, missing values in the data are put last; if `FALSE`, they are put first; if `NA`, they are removed.
`FUN`	a function used to determine which row to include, when there are multiple rows with the same 'column' value. eg `mean, min, max, var, sd, mad, prod`
`column`	the column name within `x` which contains the keys. see details
`...`	arguments passed to FUN

given an object with data, and 1 'key' column, containing possibly non-unique identifiers, create a result, which has unique values for those keys. The 'key' column could be a gene symbol, and each row could be a probe; the goal is to convert a 1-row-per-probe table into 1-row-per-gene.

something

The column should be the name of a column found in the fData(x) slot. hint: colnames(fData(x)), or fvarLabels(x)

Often some probes don't have a genesymbol, and should thus have an NA in this column. na.list controls what to do with the gene-less probes. na.last=TRUE keeps the probes & moves them to the bottom of the result. na.last=FALSE keeps the probes & leaves them in the sorted order specified by FUN. Note this is different to other implementations of na.last. na.last=NA (default) discards these probes from the result.

Mark Cowley

## Not run: 
 load("Rmisc/x.averaged.RDa.gz")
	tmp <- collapse(x.averaged, FUN=var, decreasing=TRUE, na.last=FALSE, "SymbolReannotated")

## End(Not run)

if( require(lumi) ) {
  data(example.lumi)
  ex <- example.lumi[1:20, ]
  featureData(ex)$GeneSymbol <- c("TP53", "BRCA1", "BRCA2", "KRAS", "SMAD4")
  collapsed <- collapse(ex, TRUE, NA, mean, "GeneSymbol")
  collapsed
}