R help - Oct 2009 - Trying get away from the "for" loop

Argh...

I just realized the inefficiency of one of my loops so I am trying to get rid of
it.

I have two data frames:

names(SubdivisionHouses) 
"BuildYear", "SqrFootage", "Exterior"

names(BuildingCodes)
"Year", "Codes"

I am trying to add on the the "Codes" column to according to the
"BuildYear"

Currently I am doing the following
SubdivisionHouseMod<-NULL
for(hh in 1:unique(BuildingCodes$Year))
{  
      SubdivisionHouses_tmp<-subset(SubdivisionHouses,
SubdivisionHouses$BuildYear == BuildingCodes$Year)

      BuildingCodes_tmp<-subset(BuildingCodes, BuildingCodes$BuildYear ==
BuildingCodes[hh, ]$BuildYear)

      SubdivisionHouseMod_tmp<-cbind(SubdivisionHouses_tmp,
BuildingCodes_tmp$Codes)
      SubdivisionHouseMod<-rbind(SubdivisionHouseMod,
SubdivisionHouseMod_tmp)
}

Doing this seems to work, but it also seems to be really inefficient. 

Are there any suggestions of ways to cut this down and speed things up?

Thanks again.

You do not offer a workable example, but it appears you might be doing  
a merge operation.

?merge

merge(SubdivisionHouses, BuildingCodes, by.x="BuildYear",
by.y="Year")

-- 
David

On Oct 15, 2009, at 5:27 PM, Jason Rupert wrote:
> Argh...
>
> I just realized the inefficiency of one of my loops so I am trying  
> to get rid of it.
>
> I have two data frames:
>
> names(SubdivisionHouses)
> "BuildYear", "SqrFootage", "Exterior"
>
> names(BuildingCodes)
> "Year", "Codes"
>
> I am trying to add on the the "Codes" column to according to the
> "BuildYear"
>
> Currently I am doing the following
> SubdivisionHouseMod<-NULL
> for(hh in 1:unique(BuildingCodes$Year))
> {
>      SubdivisionHouses_tmp<-subset(SubdivisionHouses,  
> SubdivisionHouses$BuildYear == BuildingCodes$Year)
>
>      BuildingCodes_tmp<-subset(BuildingCodes, BuildingCodes 
> $BuildYear == BuildingCodes[hh, ]$BuildYear)
>
>      SubdivisionHouseMod_tmp<-cbind(SubdivisionHouses_tmp,  
> BuildingCodes_tmp$Codes)
>      SubdivisionHouseMod<-rbind(SubdivisionHouseMod,  
> SubdivisionHouseMod_tmp)
> }
>
> Doing this seems to work, but it also seems to be really inefficient.
>
> Are there any suggestions of ways to cut this down and speed things  
> up?
>
> Thanks again.
-- 

David Winsemius, MD
Heritage Laboratories
West Hartford, CT

Thank you again for the suggestions to use merge:
merge(SubdivisionHouses, BuildingCodes, by.x="BuildYear",
by.y="Year")


It appears to work perfectly for the example shown below:
BuildYear<-c(1980, 1985, 1975, 1990, 1980)
SqrFootage<-c(1500, 1650, 1500, 2000, 1450)
Exterior<-c("Brick", "Stone", "Siding",
"Brick", "Siding")

SubdivisionHouses<-data.frame(BuildYear, SqrFootage, Exterior)

Year<-c(1980, 1985, 1975, 1990, 1977, 1986)
Codes<-c("G1", "G5", "G6", "B5",
"B6", "B7")
BuildingCodes<-data.frame(Year, Codes)


SubdivisionHouseMod<-NULL
for(hh in 1:unique(BuildingCodes$Year))
{  
      SubdivisionHouses_tmp<-subset(SubdivisionHouses,
SubdivisionHouses$BuildYear == BuildingCodes[hh, ]$Year)
        
      if(length(SubdivisionHouses_tmp$BuildYear) >0)
      {
              BuildingCodes_tmp<-subset(BuildingCodes, BuildingCodes$Year ==
BuildingCodes[hh, ]$Year)

              SubdivisionHouseMod_tmp<-cbind(SubdivisionHouses_tmp,
BuildingCodes_tmp$Codes)
              SubdivisionHouseMod<-rbind(SubdivisionHouseMod,
SubdivisionHouseMod_tmp)
      }
}


Thank you again. 



--- On Thu, 10/15/09, David Winsemius <dwinsemius at comcast.net> wrote:
> From: David Winsemius <dwinsemius at comcast.net>
> Subject: Re: [R] Trying get away from the "for" loop
> To: "Jason Rupert" <jasonkrupert at yahoo.com>
> Cc: R-help at r-project.org
> Date: Thursday, October 15, 2009, 4:46 PM
> You do not offer a workable example,
> but it appears you might be doing a merge operation.
> 
> ?merge
> 
> merge(SubdivisionHouses, BuildingCodes, by.x="BuildYear",
> by.y="Year")
> 
> --David
> 
> On Oct 15, 2009, at 5:27 PM, Jason Rupert wrote:
> 
> > Argh...
> > 
> > I just realized the inefficiency of one of my loops so
> I am trying to get rid of it.
> > 
> > I have two data frames:
> > 
> > names(SubdivisionHouses)
> > "BuildYear", "SqrFootage", "Exterior"
> > 
> > names(BuildingCodes)
> > "Year", "Codes"
> > 
> > I am trying to add on the the "Codes" column to
> according to the "BuildYear"
> > 
> > Currently I am doing the following
> > SubdivisionHouseMod<-NULL
> > for(hh in 1:unique(BuildingCodes$Year))
> > {
> >? ? ?
> SubdivisionHouses_tmp<-subset(SubdivisionHouses,
> SubdivisionHouses$BuildYear == BuildingCodes$Year)
> > 
> >? ? ?
> BuildingCodes_tmp<-subset(BuildingCodes,
> BuildingCodes$BuildYear == BuildingCodes[hh, ]$BuildYear)
> > 
> >? ? ?
> SubdivisionHouseMod_tmp<-cbind(SubdivisionHouses_tmp,
> BuildingCodes_tmp$Codes)
> >? ? ?
> SubdivisionHouseMod<-rbind(SubdivisionHouseMod,
> SubdivisionHouseMod_tmp)
> > }
> > 
> > Doing this seems to work, but it also seems to be
> really inefficient.
> > 
> > Are there any suggestions of ways to cut this down and
> speed things up?
> > 
> > Thanks again.
> --
> David Winsemius, MD
> Heritage Laboratories
> West Hartford, CT
> 
>