Calculate Poisson probability of one or more events — risk_calc_poisson

Calculates the Poisson probability of observing one or more events based on observed event counts and an analysis period.

Usage

risk_calc_poisson_probability(
  data,
  count_col = "event_count",
  period_col = NULL,
  period_value = 1,
  lambda_col = "lambda",
  probability_col = "prob_event_ge_1",
  probability_pct_col = "prob_event_ge_1_pct",
  output = c("proportion", "percent", "both")
)

Arguments

data: A data frame or sf object.
count_col: Name of the observed count column.
period_col: Optional column containing analysis periods.
period_value: Fixed analysis period used when period_col is NULL.
lambda_col: Name of the output lambda column.
probability_col: Name of the output probability proportion column.
probability_pct_col: Name of the output probability percent column.
output: One of "proportion", "percent", or "both".

Value

Input data with Poisson probability outputs added.

Details

This implementation uses observed historical event frequency to derive a Poisson lambda value:

$$ P(X \geq 1) = 1 - e^{-\lambda} $$

where $\lambda$ is the average event count for the target period.

References

Standard Poisson probability relationships commonly used in epidemiological and event-frequency modelling.

#' @examples data <- data.frame( event_count = c(1, 5, 10), years = c(1, 2, 5) )

risk_calc_poisson_probability( data = data, count_col = "event_count", period_col = "years", output = "both" )