One way to do this is to use rxSummary() to compute the group sums and then calculate the group percentages from these categorical counts. Here is a simple example that shows how to do this:
df <- data.frame(g = c('1', '0', '1', '0'), a=c(1, 2, 3, 4)) sums.within.g <- rxSummary(a ~ g, data = df, summaryStats = "Sum") sums.within.g <- sums.within.g$categorical[[1]][,3]numRows <- length(unique(df$g)) totSum <- cumsum(sums.within.g) totSum <- totSum[length(totSum)] df.percentages <- sums.within.g / totSum names(df.percentages) <- levels(df$g)> df.percentages  0 1 0.6 0.4
