RW2: Second-Order Random Walk Prior
In bage: Bayesian Estimation and Forecasting of Age-Specific Rates

View source: R/bage_prior-constructors.R

RW2	R Documentation

Second-Order Random Walk Prior

Description

Use a second-order random walk as a model for a main effect, or use multiple second-order random walks as a model for an interaction. A second-order random walk is a random walk with drift where the drift term varies. It is typically used with terms that involve age or time, where there are sustained trends upward or downward.

Usage

RW2(s = 1, sd = 1, sd_slope = 1, along = NULL, con = c("none", "by"))

Arguments

`s`	Scale for the prior for the innovations. Default is `1`.
`sd`	Standard deviation of initial value. Default is `1`. Can be `0`.
`sd_slope`	Standard deviation of initial slope. Default is `1`.
`along`	Name of the variable to be used as the 'along' variable. Only used with interactions.
`con`	Constraints on parameters. Current choices are `"none"` and `"by"`. Default is `"none"`. See below for details.

Details

If RW2() is used with an interaction, a separate random walk is constructed within each combination of the 'by' variables.

Argument s controls the size of innovations. Smaller values for s tend to give smoother series.

Argument sd controls variance in initial values. Setting sd to 0 fixes initial values at 0.

Argument sd_slope controls variance in the initial slope.

Value

An object of class "bage_prior_rw2random" or "bage_prior_rw2zero".

Mathematical details

When RW2() is used with a main effect,

\beta_1 \sim \text{N}(0, \mathtt{sd}^2)

\beta_2 \sim \text{N}(\beta_1, \mathtt{sd\_slope}^2)

\beta_j \sim \text{N}(2 \beta_{j-1} - \beta_{j-2}, \tau^2), \quad j = 2, \cdots, J

and when it is used with an interaction,

\beta_{u,1} \sim \text{N}(0, \mathtt{sd}^2)

\beta_{u,2} \sim \text{N}(\beta_{u,1}, \mathtt{sd\_slope}^2)

\beta_{u,v} \sim \text{N}(2\beta_{u,v-1} - \beta_{u,v-2}, \tau^2), \quad v = 3, \cdots, V

where

\pmb{\beta} is the main effect or interaction;
j denotes position within the main effect;
v denotes position within the 'along' variable of the interaction; and
u denotes position within the 'by' variable(s) of the interaction.

Parameter \tau has a half-normal prior

\tau \sim \text{N}^+(0, \mathtt{s}^2)

Constraints

With some combinations of terms and priors, the values of the intercept, main effects, and interactions are are only weakly identified. For instance, it may be possible to increase the value of the intercept and reduce the value of the remaining terms in the model with no effect on predicted rates and only a tiny effect on prior probabilities. This weak identifiability is typically harmless. However, in some applications, such as forecasting, or when trying to obtain interpretable values for main effects and interactions, it can be helpful to increase identifiability through the use of constraints.

Current options for constraints are:

"none" No constraints. The default.
"by" Only used in interaction terms that include 'along' and 'by' dimensions. Within each value of the 'along' dimension, terms across each 'by' dimension are constrained to sum to 0.