---
title: "Running a national level multi-country model"
output: 
  rmarkdown::html_vignette:
       toc: true
vignette: >
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteIndexEntry{Running a national level multi-country model}
  %\usepackage[UTF-8]{inputenc}
---


```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```


## Load your library

```{r, include=TRUE, message=FALSE, eval=FALSE}
library(mcmsupply)
library(dplyr)
set.seed(1209)
```


## Load the data 

```{r, include=TRUE, message=FALSE, eval=FALSE}
cleaned_natdata <- get_data(national=TRUE)
```


## Get the JAGS model inputs and the cleaned data

```{r, include=TRUE, message=FALSE, warning=FALSE, eval=FALSE}
pkg_data <- get_modelinputs(startyear=1990, endyear=2025.5, 
                            nsegments=12, raw_data = cleaned_natdata)

```


## Run JAGS model and get posterior point estimates with uncertainty. 

For speed and illustration purposes, we will use 10 iterations, with no burn in period and taking every third sample. This leaves only 9 samples. We DO NOT recommend this setting. The recommended settings are 80000 iterations, with 10000 burn in period and taking every 35th sample. This is commented out and listed underneath the below R code. 

```{r, include=TRUE, message=FALSE, eval=FALSE}
mod <- run_jags_model(jagsdata = pkg_data, jagsparams = NULL,
                      n_iter = 5, n_burnin = 1, n_thin = 1)
                     # n_iter = 80000, n_burnin = 10000, n_thin = 35)
```


## Check the model diagnostics 

We use this to evaluate the convergence of the model parameters. We should expect to see R-hat values of approximately 1.05. The plot function will give you a visual summary for each parameter monitored. 

```{r, include=TRUE, message=FALSE, eval=FALSE}
plot(mod$JAGS)

print(mod$JAGS)
```

Using the ggplot2 and tidybayes R packages, we will check the trace plots to assess the convergence of individual parameters.
We expect to see a 'caterpillar' like appearance of the chains over the iterations. 

```{r, include=TRUE, message=FALSE, eval=FALSE}
sample_draws <- tidybayes::tidy_draws(mod$JAGS$BUGSoutput$sims.matrix)

var <- sample_draws %>% dplyr::select(.chain, .iteration, .draw,`P[1,2,1,1]`) %>%
  dplyr::mutate(chain = rep(1:2, each=mod$JAGS$BUGSoutput$n.keep)) %>%
  dplyr::mutate(iteration = rep(1:mod$JAGS$BUGSoutput$n.keep, 2))

ggplot2::ggplot(data=var) +
  ggplot2::geom_line(ggplot2::aes(x=iteration, y=`P[1,2,1,1]`, color=as.factor(chain)))
```


## Plot posterior point estimates with uncertainty

```{r, include=TRUE, message=FALSE, eval=FALSE}
plots <- plot_estimates(jagsdata = pkg_data, model_output = mod) 
```

## Pull out estimates for a particular country and year that you are particularly interested in

```{r, include=TRUE, message=FALSE, eval=FALSE}
estimates_2018 <- pull_estimates(model_output = mod, country = 'Nepal', year=2018)

head(estimates_2018)
```


# Review the complete posterior sample of estimated method-supply shares

This function will allow you to pull out the posterior sample of estimated method supply shares. The posterior sample will be of size 'nposterior'. Note that 'nposterior' should not be larger than your total iterations (given in 'run_jags_model') 
In this example, we supply the JAGS model object and the JAGS input data to the function, we set 'nposterior=4' to pull out 4 posterior samples.

```{r, include=TRUE, message=FALSE, eval=FALSE}
post_samps <- get_posterior_P_samps(jagsdata = pkg_data, model_output = mod, nposterior=4)

head(post_samps)
```