data:
> x
M D P
1 1 25 45
2 1 26 50
3 1 29 48
4 1 30 49
5 2 29 50
6 2 30 52
7 2 31 51
a quick and dirty solution:
> rev(unlist(rev((tapply(rev(x$P),rev(x$M),FUN=function(x)cummin(x))))))
14 13 12 11 23 22 21
45 48 48 49 50 51 51
good luck for further eval
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Peter Malewski Center for Psychotherapy Research
Psychologist Christian-Belser-Strasse 79a
Project TR-EAT D-70597 Stuttgart Germany
malewski@psyres-stuttgart.de
+49 711 6781-415 (phone)
+49 711 6876902 (fax)
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
On Fri, 19 Jun 1998 andrey@utstat.toronto.edu wrote:
>
> >
> >In the example below I would like to calculate for each Day within each
> >Month the minimum Price that occurred on or after that Day within that
> >Month. A loopless solution, anyone?
> >
> > | DESIRED RESULT
> >Month Day Price | MinPrice
> > |
> > 1 25 45 | 45
> > 1 26 50 | 48
> > 1 29 48 | 48
> > 1 30 49 | 49
> >
> > 2 29 50 | 50
> > 2 30 52 | 51
> > 2 31 51 | 51
> >
> >Allen Humbolt
> >
>
>
> this problem was just too nice for me to resist trying!
> here is one, admittedly somewhat hasty, basic idea.
> say m, d, and p are the names of the three vectors (say) that you
> mention. suppose also that the months are labelled in the manner:
> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15,...
> (if they are labelled instead as:
> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 1, 2, 3,...
> the idea below can be adjusted by incorporating a dates variable also...)
> suppose also, for example, that 10^5 is larger than any of the prices.
>
>
> ## add `factors' assuring no LATER month has lower prices:
> x _ p + m*10^5
> ## next, append 31 large numbers to the end of this vector:
> x _ c ( x, rep( (max(m)+1)*10^5, 31) )
> n _ length(x)
>
> ## now, a somewhat long (but finite!) line of code:
> x _ apply (
> rbind ( x[1:(n-31)], x[2:(n-30)], x[3:(n-29)], ...etc..., x[31:(n-1)] ),
> 1,
> min )
>
> x _ x - m*10^5 ## now get rid of the added `factors'
>
> x should contain what you want.
>
> i have not had time to check this code (sorry!!), so there are likely
> to be small mistakes in it, but believe that the basic idea should work...
>
> andrey feuerverger
> university of toronto
>
> -----------------------------------------------------------------------
> This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
> send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
> message: unsubscribe s-news
>
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news
|