Calculate the Time at a Given Percentile of a Weighted Distribution

Computes the interpolated time at which a specified percentile of the cumulative sum of a weight column (e.g., waiting list size) is reached, based on a sorted data frame.

Usage

hist_percentile_calc(
  wl_structure,
  percentile = 0.92,
  wlsize_col = "wlsize",
  time_col = "months_waited_id"
)

Arguments

wl_structure

A data frame containing at least the columns specified by wlsize_col and time_col.

percentile

Numeric value between 0 and 1 indicating the desired percentile (default is 0.92).

wlsize_col

Character string specifying the name of the column representing the weights (default is wlsize).

time_col

Character string specifying the name of the column representing the time variable (default is time).

Value

A numeric value representing the interpolated time at which the cumulative sum of the weights reaches the specified percentile. The value is the number of months waited where the percentile value is met. Despite the requirement that the time_col starts at 0, a value of 0.5 means the percentile value is met after waiting 0.5 months

Details

The function sorts the data frame by the time column, computes the cumulative sum of the weights, and determines the time at which the cumulative sum first meets or exceeds the specified percentile of the total sum. If the percentile falls within a bin, linear interpolation is used to estimate the precise time.

Examples

if (FALSE) { # \dontrun{
wl_structure <- data.frame(time = 1:10, wlsize = c(5, 3, 8, 2, 7, 4, 6, 1, 9, 2))
hist_percentile_calc(wl_structure, percentile = 0.9)
} # }