R help - May 2009 - code optimization problem ... using or not using "which" function

hello all,

I have two data sets that share certain fields of of interest (
facility, unit, date) which I want to match up, and from this extract
information from one dataset and store it in the other.

my first initial idea  (which I know is bad) goes like this:

##  capacity  and new_trayloc are datasets in example code:

for( i in 1: nrow( new_trayloc) {


theshifts<-which(as.Date(capacity$shift_dt) == new_trayloc$admit_dt[i] &
      as.character(capacity$unit)==as.character(new_trayloc$UNIT_1[i]) &
      as.character(capacity$fac_id)==as.character(new_trayloc$ORIG_FAC_ID[i]))


thenightshifts<-which(as.Date(capacity$shift_dt) == new_trayloc$admit_dt[i]-1
&
      as.character(capacity$unit)==as.character(new_trayloc$UNIT_1[i]) &
      as.character(capacity$fac_id)==as.character(new_trayloc$ORIG_FAC_ID[i]))


..... obtain information by using theshifts and thenightshifts objects
and store in new_trayloc

}

. by doing a system.time on the entire for loop for 5 iterations, i
get a time of
 user  system elapsed
  25.66    1.04   26.72

That seems really bad... and plus, i need to run it for over 100,000 iterations.

Any suggestions in either the way I match the fields, or my approach
to my problem?


Cheers,
Juan Carlos

For a start, do all your conversions to character and Date once outside the
loop so you are not doing them for each iteration.  Not exactly sure what
you are doing, but it looks like with the 'and's you are only checking
for
the rows that are the same.  You might want to use a 'match' function
like:

x <- match(capacity$shift_dt, new_trayloc$admin)

to get where each of the items match and then when you have done it for the
three conditions, you then find columns that have the same number indicating
all condition match for that row.

On Fri, May 29, 2009 at 7:17 PM, Juan Carlos Laguardia <
brassman785@gmail.com> wrote:
> hello all,
>
> I have two data sets that share certain fields of of interest (
> facility, unit, date) which I want to match up, and from this extract
> information from one dataset and store it in the other.
>
> my first initial idea  (which I know is bad) goes like this:
>
> ##  capacity  and new_trayloc are datasets in example code:
>
> for( i in 1: nrow( new_trayloc) {
>
>
> theshifts<-which(as.Date(capacity$shift_dt) == new_trayloc$admit_dt[i]
&
>      as.character(capacity$unit)==as.character(new_trayloc$UNIT_1[i]) &
>
>  as.character(capacity$fac_id)==as.character(new_trayloc$ORIG_FAC_ID[i]))
>
>
> thenightshifts<-which(as.Date(capacity$shift_dt) =>
new_trayloc$admit_dt[i]-1 &
>      as.character(capacity$unit)==as.character(new_trayloc$UNIT_1[i]) &
>
>  as.character(capacity$fac_id)==as.character(new_trayloc$ORIG_FAC_ID[i]))
>
>
> ..... obtain information by using theshifts and thenightshifts objects
> and store in new_trayloc
>
> }
>
> . by doing a system.time on the entire for loop for 5 iterations, i
> get a time of
>  user  system elapsed
>  25.66    1.04   26.72
>
> That seems really bad... and plus, i need to run it for over 100,000
> iterations.
>
> Any suggestions in either the way I match the fields, or my approach
> to my problem?
>
>
> Cheers,
> Juan Carlos
>
> ______________________________________________
> R-help@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
>
http://www.R-project.org/posting-guide.html<http://www.r-project.org/posting-guide.html>
> and provide commented, minimal, self-contained, reproducible code.
>


-- 
Jim Holtman
Cincinnati, OH
+1 513 646 9390

What is the problem that you are trying to solve?

	[[alternative HTML version deleted]]