\renewcommand\thefigure{D\arabic{figure}}
\renewcommand\thetable{D\arabic{table}}

Appendix D: Model convergence diagnostics

Scale reduction factors [@Gelman1992; @Gelman2009] were less than 1.1 for all fitted parameters (Figure D1). Visual inspection of traceplots indicated convergence.

\footnotesize

Table: Definition of parameters and model notation for Figure D1.

| Parameter | Definition | Notation in figures | | --------- | ---------- | -------------------------------------- | | $\alpha_{i}$ | spatial random effect for for knot i | alpha.i where i is the numeric knot id | | $\gamma_{x}$ | climate effect for covariate x | beta.x where x is the numeric climate covariate id | | $\beta_{0t}$ | random effect for year t on the intercept | int_yr.t where t is the numeric year id | | $\bar{\beta_{0}}$ | mean intercept | int_mu | | $\beta_{1}$ | density-dependent effect | beta_mu | | $\sigma_{\beta_{0}}^2$ | variance of year random effect | sig_yr | | $\sigma_{\eta}^2$ | variance of spatial random effect | sig_a |

\normalsize

Scale reduction factors for each parameter ($\hat{R}$, points) and the unacceptable level of 1.1 shown in red.

References



atredennick/sageAbundance documentation built on May 10, 2019, 2:11 p.m.