"R provides Type I sequential SS, not the default Type III marginal SS reported by SAS and SPSS. In a nonorthogonal design with more than one term on the right hand side of the equation order will matter (i.e., A+B and B+A will produce different results)! We will need use the drop1( ) function to produce the familiar Type III results."
Math and biostats are the shortest. History is the longest. Nobody is suprised.
Includes a pretty interesting discussion of dealing with overdispersion in mixed-effects models (e.g., count models) by using individual-level fixed effects. It also includes a whole series of citations.
Cute way of creating web visualizations in R that don't work in my new NoScript-enabled browser.
A bunch of R code to fit your power law distributions.
I'm a huge fan of data.table and I've used it a ton in my own work in the last year.
There should be a better way of finding your R package than searching though a page of short descriptions of all 2800 package.s But there isn't.
Answer: When considering which class to use, always choose the least complex class that will support the application. That is, use Date if possible, otherwise use chron and otherwise use the POSIX classes.
It's exactly what it sounds like. Good? Slightly horrifying?
Awesome (and very short) section of debugging: 1. Be liberal with the use of print (or cat()) statements in your functions when debugging them! 2. traceback() # can see the sequence of function calls 3. options(error = dump.frames) debugger() # permits you to see the values of objects in the various nested environments of the function calls
A new R module that lets you keep your data on disk.
The description of fitting coxph to time dependent data here seems to be better than any I've found yet.
