---
title: "Reset a BCEA object"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Reset a BCEA object}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.width = 6,
  fig.height = 6
)
```

```{r setup, results='hide', message=FALSE, warning=FALSE, echo=FALSE}
library(BCEA)
library(ggplot2)
```


## Introduction
When running `BCEA`, we create objects that belong in a special `bcea` "class" 

```{r}
data(Vaccine)
m=bcea(eff,cost,ref=2,interventions=treats)
class(m)
```

We may want to run advanced analyses on this object. For instance, we may consider the "mixed analysis", which accounts for the potential market shares of the interventions being compared

```{r}
mixedAn(m)=c(.25,.75)
class(m)
```

--- in this case, we assume that intervention $t=1$ has 25% of the market and intervetion $t=2$ has the remaining 75%.

This operation changes the nature of the `m` object, to which we add relevant quantities/variables and an additional class `mixedAn`. This unlocks new plotting methods. For instance, when we do

```{r}
plot(m)
```

we get a specialised graph, describing the Expected Value of Information comparing the mixed with the optimal strategy (in which we allocate all the shares to the intervention that is most cost-effective, given current evidence). The gray area represent the loss due to "market inefficiency".

Interestingly, because we do not remove the class `bcea` from the object `m`, we can still use the original methods calling them specifically. For instance

```{r}
BCEA:::plot.bcea(m)
```

now returns the "standard" graphical summary of the basic economic evaluation performed by `bcea()`. 

In addition, methods for which a `mixedAn` option is not available (e.g. `ceac.plot`) can be used directly -- after all, `m` is still an object of the class `bcea` too and so if we try

```{r}
ceac.plot(m)
```

`R` will understand to apply the method `ceac.plot.bcea()` in the background.

Similar reasoning applies if we use the `CEriskav` class, which we can do by calling

```{r}
CEriskav(m)=c(.001,.002,.003)
class(m)
```

This adds another class to the object `m` and unlocks other `plot` methods. Now `CEriskav` becomes the default class and so 

```{r}
plot(m)
```

generates the specific graph to perform the analysis including risk aversion. But 

```{r}
BCEA:::plot.mixedAn(m)
```

applies the specific method to do the mixed analysis.

**IF** we want to return to the original status and remove the extra credentials generated when calling `mixedAn` and/or `CEriskav`, we can simply use the `reset_bcea()` function.

```{r}
m=reset_bcea(m)
class(m)
plot(m)
```


```{r echo=FALSE}
# # create output docs
# rmarkdown::render(input = "vignettes/evppi.Rmd", output_format = "pdf_document", output_dir = "vignettes")
# rmarkdown::render(input = "vignettes/evppi.Rmd", output_format = "html_document", output_dir = "vignettes")
```