# R Simple Data Analysis

Jun 27, 2013

While composing the results of our performance tests recently, I started investigating R to programmatically generate graphs. Excel was proving to be tedious, error-prone, and unstable. It was particularly frustrating to work with the JMeter response tables which have about 500,000 rows each. What immediately amazed me about R was how well it handled CSV files and modifying data sets. As an example using the following CSV data;

1
2
3
4
5

ts,t
1000,40
2000,47
3000,53
4000,35

Loading into R and converting time in milliseconds to seconds was as simple as;

1
2

response_times = read.csv("perf.csv")
response_times$ts = response_times$ts / 1000

Note I didn’t have to do an apply, loop, etc to divide all of the data in the set by 1000 (I tried the apply function initially). While I wouldn’t say R is ideal as a general purpose language it’s incredibly well suited for anything involving statistical analysis. If you’re looking at building graphs ggplot2 provides an easy way to generate graphs.

tags: [ R ]