Growing List vs Growing Queue

This article is originally published at https://statcompute.wordpress.com

### GROWING LIST ###
base_lst1 <- function(df) {
  l <- list()
  for (i in seq(nrow(df))) l[[i]] <- as.list(df[i, ])
  return(l)
}

### PRE-ALLOCATING LIST ###
base_lst2 <- function(df) {
  l <- vector(mode = "list", length = nrow(df))
  for (i in seq(nrow(df))) l[[i]] <- as.list(df[i, ])
  return(l)
}

### DEQUER PACKAGE ###
dequer_queue <- function(df) {
  q <- dequer::queue()
  for (i in seq(nrow(df))) dequer::pushback(q, as.list(df[i, ]))
  return(as.list(q))
}

### LIQUEUER PACKAGE ###
liqueuer_queue <- function(df) {
  q <- liqueueR::Queue$new()
  for (i in seq(nrow(df))) q$push(as.list(df[i, ]))
  return(q$data)
}

### COLLECTIONS PACKAGE ###
collections_queue <- function(df) {
  q <- collections::Queue$new()
  for (i in seq(nrow(df))) q$push(as.list(df[i, ]))
  return(q$as_list())
}

### RSTACKDEQUE PACKAGE ###
rstackdeque_queue <- function(df) {
  q <- rstackdeque::rpqueue()
  for (i in seq(nrow(df))) q <- rstackdeque::insert_back(q, as.list(df[i, ]))
  return(as.list(q))
}

nyc <- read.csv("Downloads/nycflights.csv")

compare <- function(ds) {
  tests <- c("dequer_queue(ds)",
             "base_lst2(ds)",
             "liqueuer_queue(ds)",
             "collections_queue(ds)",
             "rstackdeque_queue(ds)")
  for (t in tests) print(identical(base_lst1(ds), eval(parse(text = t))))
}

compare(nyc[1:10, ])
#[1] TRUE
#[1] TRUE
#[1] TRUE
#[1] TRUE
#[1] TRUE

### BENCHMARKS ###
bm <- function(ds) {
  rbenchmark::benchmark(replications = 5, order = "elapsed", relative = "elapsed",
                        columns = c("test", "replications", "elapsed", "relative"),
  "GROWING LIST"         = base_lst1(ds),
  "PRE-ALLOCATING LIST"  = base_lst2(ds),
  "DEQUER::QUEUE"        = dequer_queue(ds),
  "LIQUEUER::QUEUE"      = liqueuer_queue(ds),
  "COLLECTIONS::QUEUE"   = collections_queue(ds),
  "RSTACKDEQUE::RPQUEUE" = rstackdeque_queue(ds)
  )
}

bm(nyc[1:1000, ])
                  test replications elapsed relative
#1         GROWING LIST            5   0.808    1.000
#2  PRE-ALLOCATING LIST            5   0.839    1.038
#5   COLLECTIONS::QUEUE            5   0.842    1.042
#4      LIQUEUER::QUEUE            5   1.091    1.350
#3        DEQUER::QUEUE            5   1.375    1.702
#6 RSTACKDEQUE::RPQUEUE            5   1.901    2.353

bm(nyc[1:10000, ])
#                  test replications elapsed relative
#5   COLLECTIONS::QUEUE            5   8.175    1.000
#1         GROWING LIST            5   8.505    1.040
#2  PRE-ALLOCATING LIST            5  12.554    1.536
#4      LIQUEUER::QUEUE            5  17.325    2.119
#6 RSTACKDEQUE::RPQUEUE            5  21.785    2.665
#3        DEQUER::QUEUE            5  22.030    2.695

bm(nyc[1:20000, ])
#                  test replications elapsed relative
#5   COLLECTIONS::QUEUE            5  16.730    1.000
#2  PRE-ALLOCATING LIST            5  17.134    1.024
#1         GROWING LIST            5  17.342    1.037
#4      LIQUEUER::QUEUE            5  48.359    2.891
#6 RSTACKDEQUE::RPQUEUE            5  52.420    3.133
#3        DEQUER::QUEUE            5  79.938    4.778

bm(nyc[1:30000, ])
#                  test replications elapsed relative
#2  PRE-ALLOCATING LIST            5  24.600    1.000
#5   COLLECTIONS::QUEUE            5  24.797    1.008
#1         GROWING LIST            5  25.600    1.041
#6 RSTACKDEQUE::RPQUEUE            5  60.908    2.476
#4      LIQUEUER::QUEUE            5 102.482    4.166
#3        DEQUER::QUEUE            5 182.039    7.400

Thanks for visiting r-craft.org
This article is originally published at https://statcompute.wordpress.com
Please visit source website for post related comments.