Ali Oghabian shares some hard-earned advice:
The aggregate function can be very useful in R, allowing one to run a function (e.g. mean) within groups of rows, in each column in a matrix/data-frame and organize the results in an easy-to-read table. However, the function takes long to run for very wide matrices and data frames, where the number of the columns are large. I this post I demonstrate the issue and show a couple of nice solutions that at least for the example cuts down the time to 15% and even less, compared to the run-time of the aggregate function.
Click through for a demo. Granted, this is a matrix with 10,000 columns, so I’m not sure how this applies to narrower matrices. H/T R-Bloggers.