Empirical Distributions

by | Oct 18, 2010 | 0 comments

A constant source of nagging doubt in system design is the use of parametric probability density functions. For the most part it is better to use empirical distributions i.e. use the data itself to determine the liklihood of a given future occurrance. Of course, this...

Wrangling “R”: Rrangling!

by | Aug 23, 2010 | 0 comments

Trying to figure out an easy way to do Pearson’s Chi-Squared Test in R when verifying goodness of fit. I thought the function chisq.test() would help!Problems? There’s no way to tell chisq.test the correct degrees of freedom (“df”) – it...