Order of Normalization and splitting layer

[pandaqiuqiu]

According to the Seurat 5 tutorial, there are two different approaches:

1. Normalize first, then split layers by sample:
   obj <- NormalizeData(obj)
   obj[["RNA"]] <- split(obj[["RNA"]], f = obj$Sample)

2. Stratify by sample first, then normalize within each sample:
   obj[["RNA"]] <- split(obj[["RNA"]], f = obj$Sample)
   obj <- NormalizeData(obj)

Do these two methods make a difference? Which one is better?