nearestSamples: impute the gate (flagged as failure by external algorithm)...
In RGLab/flowIncubator: Routines that don't belong to the core flow packages yet.

impute the gate (flagged as failure by external algorithm) with refGate from nearest neighbour sample

1	nearestSamples(gs, node, failed, passed = NULL, ...)

`gs`	a `GatingSet`
`node`	a `character` or `numeric` specifing node index
`failed`	a `character` or `numeric` specifing samples that fails the gating QA
`passed`	a `character` or `numeric` specifing samples that passes QA to be served as references By default, all samples other than failed will be used. but sometime it is helpful to narrow it down to a few of really good samples.
`...`	other arguments passed to `.nearestSample` n an `integer` passed to `density` call. default is 512. bandwidth, gridsize passed to `bkde2D` call. default is c(5, 5) and c(100, 100). method density similarity calculation method. Either "ks.test" or "em". For 2d gate, "em" is the only choice. mc.cores passed to mclapply for parallel computing. Default is 1.

## Not run: 
library(flowWorkspace)
gs <- load_gs("gs-tcell/")

#Finding reference sample for: 1349_3_Tcell_A06.fcs
res <- nearestSamples(gs, node = "CD3", failed = "1349_3_Tcell_A06.fcs")
# res is a named vector stores the mapping between bad(failed) and good(reference) samples

#fix the gate for the bad samples
regateNearestSamples(gs, res, "CD3")

#run it on 2d gate and customize some parameters
#and enable parallel computing to speed it up
res <- nearestSamples(gs, node = "CD4", failed = "1349_3_Tcell_A06.fcs", gridsize = c(70, 70), mc.cores = 4)
#fix the 2d gate for the bad samples
regateNearestSamples(gs, res, "CD4")

## End(Not run)