The best thing about R is that it was written by statisticians. The worst thing about R is that it was written by statisticians.

I guess this NULL thing would be one of those cases.

NA would not be applicable, because it’s a placeholder for a missing value. In data analysis, you tend to have have lots of NA-values, and dealing with them is super common. Every function needs to be able to handle them gracefully. For instance mean(someValues, na.rm=TRUE) would be a command for calculating the mean of a particular vector while igrnoring all NA-values. Super handy. Excel handles these missing values in a very annoying manner, but that’s a topic for another rant.

NULL can be considered an error value, but obviously it’s not very helpful because it doesn’t tell you what went wrong. Obviously, R does have all sorts of error messages, but in this case it just says NULL instead. If you find that some variable has NULL in it, you can be pretty sure something went wrong and it’s most likely due to going outside the space of a particular variable. Likewise interger(0), character(0) or logical(0) are the results you get when you’re searching for something that doesn’t exist. Not really my favorite type of error.