[R] Relative subscripting
ggrothendieck at myway.com
Wed Dec 1 15:11:30 CET 2004
And, in general, using the right classes or data structures
makes modelling your problem in R much easier. It is
surprising that there was an entire discussion on this
before Brian brought up the right approach.
Prof Brian Ripley <ripley <at> stats.ox.ac.uk> writes:
: I've not seen anyone mention that this is really a time-series problem of
: lagging variables, and something like
: tmp <- ts.intersect(as.ts(return), lag(marketcap, -1))
: is a lot more intuitive, easier to generalize and keeps the timebase
: information around.
: On Wed, 1 Dec 2004 Ted.Harding <at> nessie.mcc.ac.uk wrote:
: > On 01-Dec-04 Uwe Ligges wrote:
: >> (Ted Harding) wrote:
: >>> [...]
: >>> Let
: >>> N<-length(return)
: >>> new.var <- return[2:N]*marketcap[1:(N-1)]
: >> Ted, I aggree to all of your points, but we can simplify by negative
: >> indices (and hence circumvent your note 1):
: >> return[-1] * marketcap[-N]
: >> Uwe Ligges
: > Neat footwork, Uwe!
: > (Why didn;t I think of that? No, don't answer ... )
: > Ted.
More information about the R-help