queue_step: Compute the departure times and queue lengths for a queueing...
In queuecomputer: Computationally Efficient Queue Simulation

queue_step

R Documentation

Compute the departure times and queue lengths for a queueing system from arrival and service times.

Description

Compute the departure times and queue lengths for a queueing system from arrival and service times.

Usage

queue_step(arrivals, service, servers = 1, labels = NULL)

Arguments

`arrivals`	numeric vector of non-negative arrival times
`service`	numeric vector of service times with the same ordering as arrival_df.
`servers`	a non-zero natural number, an object of class `server.stepfun` or an object of class `server.list`.
`labels`	character vector of customer labels (deprecated).

Details

If only departure times are needed, the queue function is faster.

Value

An list object of class queue_list with the following components:

departures - A vector of response times for the input of arrival times and service times.
server - A vector of server assignments for the input of arrival times and service times.
departures_df - A data frame with arrivals, service, departures, waiting, system time, and server assignments for each customer.
queuelength_df - A data frame describing the evolution of queue length over time
systemlength_df - A data frame describing the evolution of system length over time
servers_input - A copy of the server argument
state - A vector of availability times for the servers

Examples


# With two servers
set.seed(1)
n <- 100

arrivals <- cumsum(rexp(n, 3))
service <- rexp(n)


queue_obj <- queue_step(arrivals,
    service = service, servers = 2)


summary(queue_obj)
plot(queue_obj, which = 5)

# It seems like the customers have a long wait.
# Let's put two more servers on after time 20


server_list <- as.server.stepfun(c(20),c(2,4))

queue_obj2 <- queue_step(arrivals,
    service = service,
    servers = server_list)

summary(queue_obj2)
if(require(ggplot2, quietly = TRUE)){

    plot(queue_obj2, which = 5)

}

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