While looking for a tool to analyze a large number of test results, I came across R. Since I had read about it multiple times in the past, I decided to give it a shot. So far it has save me from writing a number of one-off scripts and I have barely scratched the surface. For instance, given a file containing transaction timings this produces mean, median, max and standard deviation:

cat trans_times.log | Rscript -e 'summary (as.numeric (readLines ("stdin")))'

Coincidentally, I was looking for a new Coursera course and came across the Computing for Data Analysis course being offered by Johns Hopkins University. Sometimes you just trip over exactly what you need.