Thx Eduard - it's about 5% faster your way according to rbenchmark. Hey, latency is latency....just got a need for speed...
Best, John On Wed, Aug 14, 2013 at 9:58 AM, Eduard Antonyan <[email protected]>wrote: > you could use Days %in% c(1,5,10,20,30,60,90) instead, or iff Days is the > first key J(c(1,5,10,20,30,60,90)) > > > On Wed, Aug 14, 2013 at 9:30 AM, John Kerpel <[email protected]>wrote: > >> Folks: >> >> I've been working more and more in data.table and it gets better and >> better as I learn to use it.... >> >> So my question is, in the following example, is the Where statement >> inefficient because I'm using "==" ? (I just want to use the subsets where >> days are equal to the values in the statement) >> >> Should I do this another way? (Btw, I get exactly the right answer using >> this approach...) >> >> db <- DB4[ >> >> Days==1 | Days==5 | Days==10 | Days==20 | Days==30 | Days==60 | >> Days==90, >> >> j = {list(v=m_func(x,y,z))}, >> >> by= "Date,Indicator,Days" >> >> ] >> >> If more detail is necessary, lmk. Date, Indicator, and Days are keys. >> >> Best, >> >> John >> >> _______________________________________________ >> datatable-help mailing list >> [email protected] >> >> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/datatable-help >> > >
_______________________________________________ datatable-help mailing list [email protected] https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/datatable-help
