microbiome::transform confusing transformation

[alimayy]

Hi, thanks for developing the microbiome package. Question about microbiome::transform():

When I provide the function with a phyloseq object where rows are OTUs and columns are samples, and run it with

OTU = otu_table(tax_count_38_mat, taxa_are_rows = TRUE)

# remove taxa with less than 1 count in more than 10% of the samples
min_count <- 1
perc_samples <- 0.1

tax_count_38_highab <- OTU[ rowSums(OTU > min_count) > ncol(OTU)*perc_samples, ]

compositional_abundance_table <- microbiome::transform((tax_count_38_highab), 
                                                       transform = 'compositional', 
                                                       target = 'OTU')

I see that the data is normalised column-wise, i.e. the sum of each column is 1. This is confusing, because the "target" parameter is specified as

target: Apply the transform for 'sample' or 'OTU'. Does not affect the log transform.
https://rdrr.io/github/microbiome/microbiome/man/transform.html

if the transform was applied to OTUs as specified in my command above, I don't think the column sums would necessarily sum up to 1 for each sample (that would require target = 'sample')?