Functions to benchmark for http://stackoverflow.com/questions/26160079/fast-concise-way-to-generate-ordered-frequency-count-of-unique-matrix-rows

SO26160079.R

library(data.table)
library(dplyr)

x <- read.csv(textConnection(
  '0,1,1,0
0,1,1,0
1,0,1,0
0,1,0,1
1,0,0,1'),
  header = FALSE)

set.seed(1)
x20 <- x[sample(nrow(x), 20, TRUE), ]
x1M <- x[sample(nrow(x), 1e6, TRUE), ]

funDplyr <- function(indf) {
  temp1 <- indf %>%
    mutate(rowid = sequence(nrow(indf))) %>%
    regroup(lapply(names(indf), as.symbol)) %>%
    mutate(rows = paste(rowid, collapse = ",")) %>%
    summarise(N = n(), rowid = rowid[1], rows = rows[1]) %>%
    arrange(rowid)
  temp1$rows <- strsplit(temp1$rows, ",", fixed = TRUE)
  temp1
}

funDplyr(x)
funDplyr(x20)
funDplyr(x1M)


funDt <- function(indt, addRN = TRUE, rnName = "rn") {
  if (!is.data.table(indt)) indt <- as.data.table(indt)
  if (isTRUE(addRN)) indt[, (rnName) := sequence(nrow(indt))]
  byCols <- setdiff(names(indt), rnName)
  indt[, list(
    .N, rowid = get(rnName)[1],
    rows = list(get(rnName))), 
    by = byCols]
}

funDt(x)
funDt(x20)
funDt(x1M)