score_trajectory: Trajectory Measure for Count Data

In lcrayton18/State_package_burdenandtrajectory: Used to calculate sub-state metrics relating to the COVID-19 pandemic

Description

Calculates the ratio of the most current to previous period counts. Values greater than 1 indicate growing; less than one indicate shrinking. NOTE: Counts must have the same period (e.g. both summed over 7 consecutive days.) In order to avoid extreme values and dividing by zeros, the ratio uses an offset of 0.5 for both the numerator and denominator as pseudocounts and is capped at 5 (i.e. a 400% increase).

Usage

score_trajectory(curr, prev)

Arguments

curr	vector of counts for current period (must be non-negative integer)
prev	vector of counts for previous period (must be non-negative integer)

Details

The 0.5 offset was inspired by the bayes.poisson.test function from the BayesianFirstAid packages uses a non-informative prior distribution whose mean simplifies to approximately 0.5. For more information, see the reproducibility paper on the OHI Surveillance Team's SharePoint site.

Value

a numeric vector

Examples

traj <- score_trajectory(curr = 100L, prev = 80L)

lcrayton18/State_package_burdenandtrajectory documentation built on July 19, 2020, 12:37 a.m.