Closed tingszhang closed 1 year ago
I think it could be true that you have no NA in the expression data. However, it might have NaN after z-score normalization if you have genes with no expression at all or expressing in a constant value. I suggested that you picked up genes with high variability to do this analysis, or you remove those with constant value or no expression.
yau.pam.pred <- runPAM(train.expr = fpkm, moic.res = cmoic.brca, test.expr = brca.yau$mRNA.expr) --all samples matched. --a total of 8545 genes shared and used. --log2 transformation done for training expression data. --testing expression profile seems to have been standardised (z-score or log transformation), no more action will be performed. Error in quantile.default(sd, offset.percent/100) : 'na.rm'如果设为FALSE的话不允许有遺漏值和NaN
I'm sure there's no NA.