queue_lengths: Compute queue lengths from arrival, service and departure...

In queuecomputer: Computationally Efficient Queue Simulation

Compute queue lengths from arrival, service and departure data

Description

Compute queue lengths from arrival, service and departure data

Usage

queue_lengths(arrivals, service = 0, departures, epsilon = 1e-10, ...)

Arguments

arrivals	vector of arrival times
service	vector of service times. Leave as zero if you want to compute the number of customers in the system rather than queue length.
departures	vector of departure times
epsilon	numeric small number added to departures to prevent negative queue lengths
...	additional arguments - does nothing, for compatibility

Examples

library(dplyr)
library(queuecomputer)

set.seed(1L)
n_customers <- 100

queueoutput_df <- data.frame(
  arrivals = runif(n_customers, 0, 300),
  service = rexp(n_customers)
)

queueoutput_df <- queueoutput_df %>% mutate(
  departures = queue(arrivals, service, servers = 2)
)

queue_lengths(
  queueoutput_df$arrivals,
  queueoutput_df$service,
  queueoutput_df$departures
)

# The dplyr way
queueoutput_df %>% do(
  queue_lengths(.$arrivals, .$service, .$departures))

n_customers <- 1000

queueoutput_df <- data.frame(
  arrivals = runif(n_customers, 0, 300),
  service = rexp(n_customers),
  route = sample(c("a", "b"), n_customers, TRUE)
)

server_df <- data.frame(
  route = c("a", "b"),
  servers = c(2, 3)
)

output <- queueoutput_df %>%
  left_join(server_df) %>%
  group_by(route) %>%
  mutate(
    departures = queue(arrivals, service, servers = servers[1])
  ) %>%
  do(queue_lengths(.$arrivals, .$service, .$departures))


if(require(ggplot2, quietly = TRUE)){
    ggplot(output) +
      aes(x = times, y = queuelength) + geom_step() +
      facet_grid(~route)
}

queuecomputer documentation built on Nov. 16, 2022, 1:07 a.m.