Using R for Actuarial Work

In March 2013, the Comprehensive R Archive Network (CRAN) released a major new version of R, the free software programming language and environment for statistical computing and graphics. R 3.0.0 takes better advantage of 64-bit memory and allows data structures that exceed two billion records!

R is useful for a number of tasks.

• Replicates most spreadsheet work.
• Allows you to use more advanced techniques, such as generalized linear models (GLMs).
• Allows you to explore your data.
• Offers downloadable packages for advanced reserving and graphing techniques.

How to Install R:

To download R please visit: https://cran.r-project.org. R is available to download in Windows, Mac, or Linux platform.

Additional Downloadable Packages:

Packages are optional programs that you download in addition to R. Packages deal with a wide variety of topics including advanced graphics. Some packages deal with topics specific to casualty actuaries. These include:

• Actuar—calculates limited expected value. This is similar to the techniques in the book Loss Models: From Data to Decisions by Klugman, Panjer and Willmot.
• ChainLadder—The ChainLadder package provides various statistical methods for loss reserving. The package has implementations of the Mack, Munich, Bootstrap, and multivariate chain-ladder methods, as well as the loss development factor curve-fitting methods of Dave Clark and GLM-based reserving models. The ChainLadder package was coauthored by CAS members Dan Murphy and Wayne Zhang!
• Tweedie—this package implements the Tweedie distribution, which is useful for modeling pure premiums.

Learn More:

One of the best ways to learn more about R is to attend the Introduction to R Limited Attendance Seminar on December 2-3 at the Arias Resort & Casino in Las Vegas. This is a great opportunity to bring your laptop and get quickly up to speed on the power of R.

This blog post is a synopsis of the Actuarial Review article, “Using the Open-Source Statistical Language for Actuarial Work,” written by Lee Bowron. For more details on R, read the full article.

Are you using R in your actuarial work? Please let us know by leaving a comment in the section below.