For cases with filing times arrivals and disposition times
dispositions inside a window \([0, T]\), the time-average number of
pending cases equals the filing rate times the mean time to disposition,
with no probability model at all: the integral of the pending count is the
sum of the case durations (Research.P16.occupancy_integral,
little). A court that publishes a mean disposition time and a filing
rate has therefore published its average backlog (little_backlog),
and a backlog target fixes the mean disposition time that achieves it
(little_target). Cases still pending at \(T\) make the mean of
disposed cases a truncated estimate; the function reports how many are
censored so the reader sees the dark-figure problem that follows.
Arguments
- arrivals
Filing times (non-negative).
- dispositions
Disposition times, at least the filing times;
NAfor cases still pending athorizon.- horizon
Window end \(T\); by default the latest disposition.
- target_backlog
Optional backlog target \(L^*\) for which the required mean disposition time is reported.
Value
A list with n, n_censored, horizon,
filing_rate, mean_disposition_time (over disposed cases),
average_backlog (\(\lambda \bar W\)), occupancy_integral,
pending_at_horizon, required_mean_time (when a target is
given) and theorems.
Examples
a <- c(0, 1, 2, 4, 5, 7); d <- c(3, 2.5, 6, 5, 9, 10)
morie_court_backlog(a, d, horizon = 10, target_backlog = 1)
#> $n
#> [1] 6
#>
#> $n_censored
#> [1] 0
#>
#> $horizon
#> [1] 10
#>
#> $filing_rate
#> [1] 0.6
#>
#> $mean_disposition_time
#> [1] 2.75
#>
#> $average_backlog
#> [1] 1.65
#>
#> $occupancy_integral
#> [1] 16.5
#>
#> $pending_at_horizon
#> [1] 0
#>
#> $little_identity_check
#> [1] 0
#>
#> $required_mean_time
#> [1] 1.666667
#>
#> $theorems
#> [1] "Research.P16.occupancy_integral" "Research.P16.little"
#> [3] "Research.P16.little_backlog" "Research.P16.little_target"
#>
