[R-sig-dyn-mod] Event value based on the state of the system at the time the event occurs

[Daniel Kaschek]

Hi John,

you might have a look at help(events), example 2. I think, you need an
event function. Within the event function you can access all states a
the current time point of integration.

Cheers,
Daniel

On So, 2015-10-18 at 11:23 -0700, John Harrold wrote:
> Howdy Folks,
> 
> I've run into a situation where I need to create events where the values
> are dependent on one of the current state values. To give you an example,
> consider the system described below.
> 
> the values (c(4,4)) actually need to be c(4*var[2], 4*var[2]), where var[2]
> is the value of the state var[2] at times 1 and 6, respectively. I don't
> know if what I want to do is possible using the event framework. I may have
> to write a wrapper that integrates from one event to the next and restarts
> the system (this is what I've done in Matlab), but I wanted to see if there
> was some kind of callback routine that I might be able to use so I can
> stick with the framework deSolve provides.
> 
> Thanks
> John
> 
> 
> derivs <- function(t, var, parms) {
>   list(c(-var[1]*parms[1], var[2]*(1-var[2]/parms[2])))
> }
> 
> yini <- c(v1 = 0, v2=1)
> times <- seq(0, 10, by = 0.1)
> 
> eventdat <- data.frame(var    = c("v1", "v1"),
>                        time   = c(1, 6) ,
>                        value  = c(4, 4),
>                        method = c("add", "mult", "rep", "add"))
> 
> eventdat
> parms = c(0.1, 10)
> out <- vode(func   = derivs,
>             y      = yini,
>             times  = times,
>             parms  = parms,
>             events = list(data = eventdat))
> plot(out)
> 
>