**Ben Mazzotta's Weblog » R**, and kindly contributed to R-bloggers)

What do you use for network analysis? I found the Wikipedia list of network software entirely overwhelming. I wanted to test out some of the introductory tools, but avoid the trap of sinking my time into a dead-end software project. (Remember learning Minitab in freshman statistics? How often do you use Minitab today for anything other than freshman statistics?)

UCINET came very well recommended by several friends who use social network analysis for business and politics. I wasn’t sure whether it would turn out to be a proprietary, one-way data sink like Blackboard or a useful, interoperable tool for analyzing and sharing data.

The Journal of Statistical Software published an issue about Statnet in 2008. Esteemed authors in the field wrote a number of great tutorials about Statnet, which is an umbrella package for most of R’s network analysis tools.

UC Irvine hosts a wiki on network analysis tools. That wiki describes a few of the F/OSS tools. All the links contain tutorials. You have your choice of tools that are freestanding, part of R, or part of Sage/Python. If you’ve got a good reason to go with one of the other F/OSS tools, such as Guess, Pajek or Networkx, please comment with your thoughts.

I learned a bit about networks at the Santa Fe Institute a couple of years ago. Mark Newman gave stellar lectures, but to be honest a lot of the content there was either over my head, or beyond my programming skills, or not closely related to my research interests. Newman writes his own software which is no doubt more efficient than the widely used packages.

I am unabashedly in favor of R. Given the choice, I will always try to knock out a solution in R. Not only is R free and open source, it seems to be the industry standard in statistics. It also has a famously steep learning curve and requires users to learn the command line interface. If you’re not ready to write the script out by hand, you probably need a different solution.

I created a picture of industry alliances since 2000 in the space/satellite industry yesterday, and learned a lot about the industry (not just about how to use the software) from the processs. Though they sound fuzzy at first, network measures of position, groups, and distance turn out to extremely intuitive and useful. It’s extremely valuable to have some quantitative tools to back up statements such as, “There is a loose consortium of firms that cluster around a couple of key players.” If you’re reasonably comforatable with basic R commands and have a clear idea what network analysis can tell you, Statnet is ready for prime time.

And by the way, I forgot to mention the Robert Hanneman book, a free online textbook for social network analysis. Has anyone read this closely? It seems at first glance like a lucid introduction to a branch of mathematics that can sound a bit technical at first. Especially when mathematicians seem to agree that network analysis is NP-hard and sampling techniques are either extremely computationally intensive, or ineffective, or both.

**leave a comment**for the author, please follow the link and comment on their blog:

**Ben Mazzotta's Weblog » R**.

R-bloggers.com offers

**daily e-mail updates**about R news and tutorials on topics such as: Data science, Big Data, R jobs, visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...