I had a good time with R for statistical process control. I don't think I used any libraries at the time (this was a long time ago) except maybe Pareto charts, but I primarily worked with mean time between failures of events. My issue was I did not control the faulty applications, but received a lot of "blame" for their failures as we managed the high performance computer systems. I wanted to be able to shine a flashlight on the various operational jobs and the frequency of their failures. I generally felt the application owners were often times lazy in getting to a root cause analysis, and it was never my job to do this with their specialized applications. It did a pretty good job taking the heat off my team, the results were automatically generated and always available. But, changing the culture was always difficult.