get_runtiming_samples: Estimate parameters of hierarchical logistic regression using...

In jBernardADFG/ChenaSalchaTowerSonar: Apportion and enumerate salmon on the Chena and Salcha

Description

The sonar species apportionment model uses species priors that are weakly informed by a logistic curve. To estimate these parameters, a hierarchical logistic regression model is used where the non-expanded visual counts for each day are used as input data. Regression coefficients are treated as hierarchically distributed. Since the timing of the chum and king runs for the Chena and Salcha rivers tend to be similar each year, the coefficient pairs (intercepts for both rivers, slopes for both rivers) are assumed to be MVN, allowing the inclusion of a correlation parameter. If one river's sonar season is compromised, this allows the borrowing of information from the other river.

Usage

get_runtiming_samples(runtiming_data, file_dir, niter = 1e+05, ncores = 3)

Arguments

runtiming_data	Output of prepare_runtiming_data.
niter	The number of MCMC iterations. 100k will take roughly 6 minutes whereas 500k will take 30 minutes.
ncores	The number of cores for parallel chains.

Author(s)

Matt Tyres and Jordy Bernard.

jBernardADFG/ChenaSalchaTowerSonar documentation built on Oct. 11, 2020, 1 a.m.