In the first part of this article we built a function (rocdata) to calculate the co-ordinates for the ROC plot and its summary statistics. Now we need to actually produce the plot. I make most of my plots in ggplot2 because of it’s versatility. However there’s no reason why these plots couldn’t be produced

Read more »

I was recently asked to summarise an analysis using a ROC (Receiver-operator characteristics) plot. R has a number of particularly good tools to produce ROC plots – ROCR, pROC and the Bioconductor package ROC to name a few. However I thought it would be a useful exercise

Read more »

I do like Dropbox. I’m a big fan of easy to use solutions to common problems. For me I work at two different sites and I do a significant amount of work at my desktop machine at home as well as on my laptop. Dropbox solved the problem of synchronising files between

Read more »