guess_initial_parameters: Produce initial parameter estimates based on mortality data

In cyclomort: Survival Modeling with a Periodic Hazard Function

Description

Uses a basic flexsurvreg exponential mortality model to find the average hazard value, and fits a mixed normal distribution model to estimate the peaks, season durations, and weight distributions for the model. These estimates are not meant to be fully accurate but instead are meant to be good initial guesses for the fit_cyclomort function.

Usage

guess_initial_parameters(x, n, null_fits)

Arguments

x	cycloSurv object representing time of death or censorship
n	expected number of mortality seasons within a period
null_fits	original estimate for mortality rate assuming constant hazard function

Value

a named vector of guesses for parameter values, used to initialize the fitting process

cyclomort documentation built on Aug. 20, 2020, 5:06 p.m.