1 Software Installation

Please install the software below before the workshop so we can move quickly through the Day 1 material. If you have any questions, contact the course instructor, before the course.

1.0.1 R and RStudio

Install the latest version of R (currently R version 4.6.1 (2026-06-24)). We strongly recommend updating your R version before NME, as older versions may not be fully compatible.

We will use RStudio as our IDE for the workshop. Please install the most recent version.

1.0.2 R Packages

With an open R console, install EpiModel, statnetWeb, Rglpk, tsna, and their dependencies:

Code

install.packages(c("EpiModel", "statnetWeb", "Rglpk", "tsna"), dependencies = TRUE)

This will also install the needed statnet packages (ergm, tergm, network, networkDynamic, sna) and the network-animation package ndtv, which the visualization module uses. tsna (temporal network measures) is named explicitly because it is not otherwise pulled in.

Rglpk is listed separately because it would not otherwise be installed: it is a suggested package of ergm rather than a required one, and dependencies = TRUE picks up the suggestions of the packages you name, not those of their dependencies. ergm uses it as its linear programming solver when fitting dyad-dependent models, which is most of the models in this course. Without it, ergm falls back to a slower solver and prints a message to that effect. Everything still runs, but we recommend installing it.

Note

Rglpk builds on the GNU Linear Programming Kit. On macOS and Windows the version from CRAN already includes it, so there is nothing else to do. On Linux you may need the system library first, for example sudo apt-get install libglpk-dev on Debian or Ubuntu, before the R package will install.

1.0.3 Testing the Installation

After installation, verify that the key packages load correctly:

Code

library("EpiModel")
library("statnetWeb")
library("ergm.ego")
library("ndtv")
library("sna")
library("tsna")

Then confirm your installed versions match those listed below:

Package	Version
EpiModel	2.6.1
ergm	4.12.0
tergm	4.2.2
network	1.20.0
networkDynamic	0.12.0
ergm.ego	1.1.4
statnetWeb	0.6.1
Rglpk	0.6.5.1
ndtv	0.13.4
sna	2.8
tsna	0.3.6

Rglpk is not in the library() list above because ergm calls it internally rather than expecting you to attach it. Its appearing in the version table is confirmation enough that it is installed and available.

1.0.4 Statnet Tutorials

We encourage you to explore the Statnet Workshops tutorials to practice your R and network analysis skills before the course.

# Software Installation Please install the software below **before** the workshop so we can move quickly through the Day 1 material. If you have any questions, contact the [course instructor](mailto:samuel.m.jenness@emory.edu), before the course. ### R and RStudio Install the latest version of [R](https://cran.r-project.org/) (currently `r R.version.string`). We strongly recommend updating your R version before NME, as older versions may not be fully compatible. We will use [RStudio](https://posit.co/download/rstudio-desktop/) as our IDE for the workshop. Please install the most recent version. ### R Packages With an open R console, install `EpiModel`, `statnetWeb`, `Rglpk`, `tsna`, and their dependencies: ```{r, eval=FALSE} install.packages(c("EpiModel", "statnetWeb", "Rglpk", "tsna"), dependencies = TRUE) ``` This will also install the needed **statnet** packages (`ergm`, `tergm`, `network`, `networkDynamic`, `sna`) and the network-animation package `ndtv`, which the visualization module uses. `tsna` (temporal network measures) is named explicitly because it is not otherwise pulled in. `Rglpk` is listed separately because it would not otherwise be installed: it is a suggested package of `ergm` rather than a required one, and `dependencies = TRUE` picks up the suggestions of the packages you name, not those of their dependencies. `ergm` uses it as its linear programming solver when fitting dyad-dependent models, which is most of the models in this course. Without it, `ergm` falls back to a slower solver and prints a message to that effect. Everything still runs, but we recommend installing it. ::: {.callout-note} `Rglpk` builds on the GNU Linear Programming Kit. On macOS and Windows the version from CRAN already includes it, so there is nothing else to do. On Linux you may need the system library first, for example `sudo apt-get install libglpk-dev` on Debian or Ubuntu, before the R package will install. ::: ### Testing the Installation After installation, verify that the key packages load correctly: ```{r, message=FALSE, warning=FALSE, results='hide'} library("EpiModel") library("statnetWeb") library("ergm.ego") library("ndtv") library("sna") library("tsna") ``` Then confirm your installed versions match those listed below: ```{r, echo=FALSE} packages <- c("EpiModel", "ergm", "tergm", "network", "networkDynamic", "ergm.ego", "statnetWeb", "Rglpk", "ndtv", "sna", "tsna") versions <- sapply(packages, function(p) as.character(packageVersion(p))) knitr::kable(data.frame(Package = packages, Version = versions), row.names = FALSE) ``` `Rglpk` is not in the `library()` list above because `ergm` calls it internally rather than expecting you to attach it. Its appearing in the version table is confirmation enough that it is installed and available. ### Statnet Tutorials We encourage you to explore the [Statnet Workshops](https://statnet.org/workshops/) tutorials to practice your R and network analysis skills before the course.