Skip to content
Prev 10957 / 398502 Next

Using R for processing computer performance data

"Peter Gallanis" <peter at gallanis.com> writes:
by() isn't really designed to be used like that. First argument would
be a data frame and the function would be something working on an
entire frame. 

However, the tapply looks as if it should work (and I wouldn't be
surprised if by() was essentially equivalent to tapply when used that
way). The error you are getting is what you get for attempting to use
boxplot() on a factor variable, which would indicate that something is
wrong with your "busy" variable. Notice that read.csv will turn
variables into factors if they contain *any* values that cannot be
interpreted as numeric (or missing -- notice the na.strings argument).

Apart from that, I suspect that what you really want is

boxplot(Busy~device, data=diskact)