Troubleshooting R's rowsum

Question

I am troubleshooting the R's row sum function. I have the following vector called total:

The vector has 20 different categories, and I would like to sum all the values for each category. My code is:

 rowsum(total[,c(1:20)], group = c(1:20))

But I get the following error:

 Error in rowsum.default(total[, c(1:2)], group = c(1:20))   incorrect length for 'group'

I am a bit confused because in the documentation, "group" is a vector/factor giving the grouping with one element per row of x.

Any help would be appreciated.

Thank you!

Both arguments should be of the same length. So you need to pass the entire first column with the group ids as the second argument. — Ramnath, Mar 18 '14 at 18:58

score 2 · Accepted Answer · answered Mar 18 '14 at 18:58

2

The second argument needs to be group memberships of the first one.

dat <- data.frame(
  value = runif(100),
  group = sample(1:20, 100, replace = T)
)

rowsum(dat$value, dat$group)

answered Mar 18 '14 at 18:58

Ramnath

1

Thank you for taking the time of answering. I am sorry, but I don't understand. Could you please explain each step? I know that you are invoking a normal distribution. – Johnathan Mar 18 '14 at 20:33

score 0 · Answer 2 · answered Mar 18 '14 at 18:53

0

You can also use tapply when you would like to break up a data into groups and do the same computation in the individual groups. You can see http://www.r-bloggers.com/r-function-of-the-day-tapply/ for more information.

answered Mar 18 '14 at 18:53

Dinesh

2 Answers2