Skip to Content


Part 2 of a series
by Daniel Hanson, with contributions by Steve Su (author of the GLDEX package)  

Recap of Part 1

In our previous article, we introduced the four-parameter Generalized Lambda Distribution (GLD) and looked at fitting a 20-year set of returns from the Wilshire 5000 Index, comparing the results of two methods, namely the Method of Moments, and the Method of Maximum Likelihood.  

The ability to create reproducible research is an important topic for many users of R. So important, that several groups in the R community have tackled this problem.

A friend was extolling the benefits of real paper books the other day, so I sent him this. Apparently, IKEA thinks the same way.


That's all for this week — we've been very busy getting ready for the Strata conference in New York. Have a great weekend, and maybe see you there!

In case you missed them, here are some articles from September of particular interest to R users.

Norm Matloff argues that T-tests shouldn't be part of the Statistics curriculum and questions the "star system" for p-values in R.

by Daniel Hanson, with contributions by Steve Su (author of the GLDEX package). Part 1 of a series.

As a computer scientist, RStudio's Joe Cheng has some great insights into the R language and how it compares with other programming language. In the interview with DataScience.LA below, he notes that while R is often thought about as a domain-specific language (or DSL), the combination of a functional language with deferred evaluation of functional arguments actually makes it a great general-purpose language for implementing a statistical DSL.


I've always loved David Fincher's films, but until I saw this video about his directing style, I never knew exactly why. His films are often about topics that don't immediately interest me (serial killers, fighting), but for some reason they're always compelling. This video explains how Fincher's direction makes them so:

by Andrie deVries

One of the reasons that R is so popular is the CRAN archive of useful packages. However, with more than 5,900 packages on CRAN, many organisations need to maintain a private mirror of CRAN with only a subset of packages that are relevant to them.