shrink_var: Split datasets
In yewei369/clotu: Classification of OTUs

Description Usage Arguments Value Examples

Split dataset into training/validation/test sets

1	shrink_var(da1, da2, thr, nvar, pair = c(1, 2), method = 0)

`da1,`	normalized dataset for target 1
`da2,`	normalized dataset for target 2
`thr,`	threshold for zombie OTU
`nvar,`	number of OTUs after shrinking
`pair,`	character vector of specified target pairs, should be of length 2
`method,`	digits for shrinking method, 0 for Kaul's method, 1 for Jun Li's

a list of shrinked OTU tables for each target and the index of significant OTUs

da<-simotu.gaus(50,700,3,nref=5,full.mean=10000,unif.min=0,unif.max=0.4,seed=1234)  
al<-data_extract(da,Target %in% c("target1","target2","target3")) # no otu names
ana0<-zomotu(al,0) ## delete otus not present in any sample, no otu names
ta<-target_split(da,ana0$otu,"Target")
ta_norm<-normalizer(ta,0,1,ana0$ref,FALSE,TRUE)
shrink_var(ta_norm[[1]],ta_norm[[2]],0,75,method=1)