I have a dataset of the form
Year tosk.fai tosk.isd tosk.gr ... tosk.total hysa.fai
hysa.isd ...
and so on. I want to sum all the columns using the first four letters in
the columns label(e.g. 'tosk', 'hysa' etc.). How can you do that? Also,
the sums should be without the
Gidday,
Perhaps try something along these lines:
## Establish which 4-letter group each row belongs to
prefix - substr(names(d), 1, 4)
gp - match(prefix, unique(prefix))
gp[regexpr(\\.total$, names(d)) -1] - NA # Exclude `*.total' rows
## Sum up each of the groups
d.sums -