I have a pipeline for targeted RNASeq data that includes edgeR for initial normalization followed by RUVSeq (RUVr). Output of RUVr is expected: df$W and df$NormalizedCounts. However, the normalized counts are in logCPM and I would like to back log them for data exploration purposes. This level of stats is a bit over my head (I'm a PhD physiology student) so I am struggling to understand what direction to take my code to back log them. If I simply take the normalized counts and apply 10^(x+1) I get counts that seem unreasonably high. This makes sense to me given that "The normalized counts are indeed simply the residuals from ordinary least squares regression of logY on W (with the offset if needed)". However, I don't understand how to apply that equation with the data in R to work back to the CPM. Can someone point me to some general R code or help with the concept of what to do to get back to CPM rather than logCPM?