percentile_rank: Percentile rank

In jmbarbone/jordan: Miscellaneous, Analytic R Kernels

Percentile rank

Description

The bounds of the percentile rank are > 0 and < 1 (see Boundaries)

A percentile rank here is the proportion of scores that are less than the current score.

PR = (c_L + 0.5 f_i) / N

Where

c_L is the frequency of scores less than the score of interest

f_i is the frequency of the score of interest

Usage

percentile_rank(x, weights = times, times)

Arguments

x	A vector of values to rank
weights, times	A vector of the number of times to repeat x

Details

Computes a percentile rank for each score in a set.

Value

The percentile rank of x between 0 and 1 (see Boundaries)

Boundaries

While the percentile rank of a score in a set must be exclusively within the boundaries of 0 and 1, this function may produce a percentile rank that is exactly 0 or 1. This may occur when the number of values are so large that the value within the boundaries is too small to be differentiated.

Additionally, when using the weights parameter, if the lowest or highest number has a value of 0, the number will then have a theoretical 0 or 1, as these values are not actually within the set.

Examples

percentile_rank(0:9)
x <- c(1, 2, 1, 7, 5, NA_integer_, 7, 10)
percentile_rank(x)

if (package_available("dplyr")) {
  dplyr::percent_rank(x)
}

# with times
percentile_rank(7:1, c(1, 0, 2, 2, 3, 1, 1))

jmbarbone/jordan documentation built on April 1, 2024, 7:46 p.m.