### Install BUGSnet Package

Source: https://bugsnetsoftware.github.io/instructions.html

Install BUGSnet and its dependencies from GitHub using the remotes package.

```R
install.packages(c("remotes", "knitr"))
remotes::install_github("audrey-b/BUGSnet@v1.1.2", upgrade = TRUE, build_vignettes = TRUE, dependencies = TRUE)
```

--------------------------------

### Verify Rtools Installation

Source: https://bugsnetsoftware.github.io/instructions.html

Check if build tools are correctly configured for R package development.

```R
install.packages("pkgbuild")
pkgbuild::has_build_tools()
```

--------------------------------

### Update R on Windows

Source: https://bugsnetsoftware.github.io/instructions.html

Use the installr package to quickly update the R version on Windows systems.

```R
install.packages("installr")
installr::updateR(TRUE)
```

--------------------------------

### BUGSnet Analysis Workflow

Source: https://bugsnetsoftware.github.io/testscript.txt

A complete sequence of operations for performing a network meta-analysis using BUGSnet, from data preparation to model comparison.

```R
library(BUGSnet)

data(diabetes.sim)

## prepare the data for use in bugsnet
diabetes.data <- data.prep(arm.data = diabetes.sim,
                           varname.t = "Treatment",
                           varname.s = "Study")

## network Plot
net.plot(diabetes.data, node.scale = 3, edge.scale=1.5)

##Network characteristic summary tables
network.char <- net.tab(data = diabetes.data,
                        outcome = "diabetes",
                        N = "n",
                        type.outcome = "rate2",
                        time = "followup")

## assessing patient covariates
data.plot(data = diabetes.data,
          covariate = "age",
          half.length = "age_SD", #comment this line out to remove error bars
          by = "treatment", #change to "study" to plot characteristics by study
          fill.str = "age_type",  #comment this line out to remove colors
          avg.hline=TRUE) #add overall average line?


## choosing an NMA model: fixed vs random effects
FE_model <- nma.model(data=diabetes.data,
                      outcome="diabetes",
                      N="n",
                      reference="Diuretic",
                      family="binomial",
                      link="cloglog",
                      time = "followup",
                      effects="fixed")

RE_model <- nma.model(data=diabetes.data,
                      outcome="diabetes",
                      N="n",
                      reference="Diuretic",
                      family="binomial",
                      link="cloglog",
                      time = "followup",
                      effects= "random")

cat(RE_model$bugs)

## running the models, will take ~30-60 seconds
FE_results <- nma.run(FE_model,
                      n.adapt=101,
                      n.burnin=0,
                      n.iter=101)

RE_results <- nma.run(RE_model,
                      n.adapt=101,
                      n.burnin=0,
                      n.iter=101)

par(mar=c(1,1,1,1))
nma.diag(RE_results, plot_prompt = FALSE)

## model fit
par(mfrow = c(1,2))
fe_model_fit <- nma.fit(FE_results, main = "Fixed Effects Model")
re_model_fit <- nma.fit(RE_results, main= "Random Effects Model")


## Synthesis of results

#Forest Plot
nma.forest(RE_results,
           central.tdcy="median",
           comparator = "Placebo", #change this option any treatment of interest, e.g "ARB"
           log.scale = FALSE)

# SUCRA Plot
sucra.out <- nma.rank(RE_results, largerbetter=FALSE, sucra.palette= "Set1")

# League table
league.out <- nma.league(RE_results,
                         central.tdcy="median",
                         order = sucra.out$order,
                         log.scale = FALSE,
                         low.colour = "springgreen4",
                         mid.colour = "white",
                         high.colour = "red")

## Run inconsistency model
re_inconsistency_model <- nma.model(data=diabetes.data,
                                   outcome="diabetes",
                                   N="n",
                                   reference="Diuretic",
                                   family="binomial",
                                   link="cloglog",
                                   time = "followup",
                                   type = "inconsistency", #specifies inconsistency model
                                   effects="random")

re_inconsistency_results <- nma.run(re_inconsistency_model,
                                   n.adapt=101,
                                   n.burnin=0,
                                   n.iter=101)

re_model_fit <- nma.fit(RE_results, main = "Consistency Model" )
inconsist_model_fit <- nma.fit(re_inconsistency_results, main= "Inconsistency Model")

graphics.off()
nma.compare(re_model_fit, inconsist_model_fit)
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.