modeled-event-time.Rmd

---
title: "Modeled events"
author: ""
date: "`r Sys.Date()`"
output: 
  rmarkdown::html_vignette: 
    toc: true
    number_sections: true
    keep_md: true
    fig_width: 5
    fig_height: 4
vignette: >
  %\VignetteIndexEntry{Modeled events}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r setup, include = FALSE}
source("global.R")
set.seed(23421)
```

# Introduction
A modeled event is a discontinuity introduced into the simulation at a time
and with characteristics that are determined within the model itself, rather
than the standard setup of specifying these events via the data set.  

In this vignette, we'll focus on creating a discontinuity in the simulation so 
that the value of a parameter can change at a specific time that is not 
necessarily chosen prior to the run time. 


# Example with single mtime
For example, we might wish a rate constant to change at some time 
that is not found in a record in the input data set.  We can make this happen
from the model code itself using `mtime()`. 

Let's make KA change at 2.1 hours (`change_t` parameter in the example below).
```{r, eval=FALSE, code = readLines("mtime-model.txt")}

```

Again, the main motivation for this is just convenience and economy of code:
we register the event time and get that time returned into a variable that 
we can reference later on, checking if we are past that time or not.


```{r, message = FALSE, warning = FALSE}
library(mrgsolve)
library(dplyr)

mod <- mread_cache("mtime-model.txt")

mod %>% ev(amt=100) %>% mrgsim(delta = 0.222) %>% plot(CENT+KA~time)
```

You won't see the message that we actually stumbled on `2.1` hours in the 
simulation even though it was not in the lineup when the simulation started. 

# Example with several mtimes

We could keep track of several mtimes like this

```{r,eval=FALSE,code = readLines("mtime-model-2.txt")}

```

```{r, message = FALSE}
mod <- mread_cache("mtime-model-2.txt")

mod %>% ev(amt=100) %>% mrgsim() %>% plot(CENT + KA ~ time)
```