Defines a list specifying the optional arguments for the time-varying reproduction number. Custom settings can be supplied which override the defaults.

## Arguments

- prior
List containing named numeric elements "mean" and "sd". The mean and standard deviation of the log normal Rt prior. Defaults to mean of 1 and standard deviation of 1.

- use_rt
Logical, defaults to

`TRUE`

. Should Rt be used to generate infections and hence reported cases.- rw
Numeric step size of the random walk, defaults to 0. To specify a weekly random walk set

`rw = 7`

. For more custom break point settings consider passing in a`breakpoints`

variable as outlined in the next section.- use_breakpoints
Logical, defaults to

`TRUE`

. Should break points be used if present as a`breakpoint`

variable in the input data. Break points should be defined as 1 if present and otherwise 0. By default breakpoints are fit jointly with a global non-parametric effect and so represent a conservative estimate of break point changes (alter this by setting`gp = NULL`

).- future
A character string or integer. This argument indicates how to set future Rt values. Supported options are to project using the Rt model ("project"), to use the latest estimate based on partial data ("latest"), to use the latest estimate based on data that is over 50% complete ("estimate"). If an integer is supplied then the Rt estimate from this many days into the future (or past if negative) past will be used forwards in time.

- gp_on
Character string, defaulting to "R_t-1". Indicates how the Gaussian process, if in use, should be applied to Rt. Currently supported options are applying the Gaussian process to the last estimated Rt (i.e Rt = Rt-1 * GP), and applying the Gaussian process to a global mean (i.e Rt = R0 * GP). Both should produced comparable results when data is not sparse but the method relying on a global mean will revert to this for real time estimates, which may not be desirable.

- pop
Integer, defaults to 0. Susceptible population initially present. Used to adjust Rt estimates when otherwise fixed based on the proportion of the population that is susceptible. When set to 0 no population adjustment is done.

## Examples

```
# default settings
rt_opts()
#> $prior
#> $prior$mean
#> [1] 1
#>
#> $prior$sd
#> [1] 1
#>
#>
#> $use_rt
#> [1] TRUE
#>
#> $rw
#> [1] 0
#>
#> $use_breakpoints
#> [1] TRUE
#>
#> $future
#> [1] "latest"
#>
#> $pop
#> [1] 0
#>
#> $gp_on
#> [1] "R_t-1"
#>
#> attr(,"class")
#> [1] "rt_opts" "list"
# add a custom length scale
rt_opts(prior = list(mean = 2, sd = 1))
#> $prior
#> $prior$mean
#> [1] 2
#>
#> $prior$sd
#> [1] 1
#>
#>
#> $use_rt
#> [1] TRUE
#>
#> $rw
#> [1] 0
#>
#> $use_breakpoints
#> [1] TRUE
#>
#> $future
#> [1] "latest"
#>
#> $pop
#> [1] 0
#>
#> $gp_on
#> [1] "R_t-1"
#>
#> attr(,"class")
#> [1] "rt_opts" "list"
# add a weekly random walk
rt_opts(rw = 7)
#> $prior
#> $prior$mean
#> [1] 1
#>
#> $prior$sd
#> [1] 1
#>
#>
#> $use_rt
#> [1] TRUE
#>
#> $rw
#> [1] 7
#>
#> $use_breakpoints
#> [1] TRUE
#>
#> $future
#> [1] "latest"
#>
#> $pop
#> [1] 0
#>
#> $gp_on
#> [1] "R_t-1"
#>
#> attr(,"class")
#> [1] "rt_opts" "list"
```