Each row of my dataframe has these data items:
Stuff1 Stuff2 TripID StopID moreStuff1 moreStuff2 ...
I might have 2 entries for TripID=9011890 and StopID=Reseda, while I
know the "Universe" for that combination was 7. I'd like to set a new
variable (Call it X1) with a value of 3.5 for the two entries with
TripID=9011890 and StopID=Reseda.
It gets messy when I have 0 entries for a combination. My "expanded"
TripID/StopID X1 values don't add to the total for the TripID. If I sum
"X1" for TripID=9011890 I'll get only say 23 when the sum of the targets
was 34. I want to have another variable (call it X2) with a value
34/23.
I've started by making a .csv file that's a "table" with each row
listing a TripID and each column a StopID. I cannot combine the .csv
file data with the dataframe data. Even though I'm using the same
commands for each, I get different "data structure" results.
My current script is as follows:
####
SurveyData <- read.spss( {Snip} )
NewTargetData <- read.table( {Snip} )
#
CurrentX1Sums <- as.matrix(xtabs(~TripID+StopID, data=SurveyData))
CurrentX2Sums <- apply(CurrentX1Sums, 1, sum)
NewTargetX1Sums <- as.matrix(NewTargetData)
NewTargetX2Sums <- apply(NewTargetX1Sums, 1, sum)
Robert Farley
Metro
www.Metro.net
-----Original Message-----
From: Erik Iverson [mailto:[EMAIL PROTECTED]
Sent: Wednesday, October 29, 2008 19:58
To: Farley, Robert
Subject: Re: [R] Trying to "expand" some data - Newbie needs help
Farley, Robert wrote:
> I want to calculate "expansion factors" for elements in my dataframe
> based on a 2-d cross classification. Since I'll have "missing values"
> (many combinations will have no record) I'll need a second "expansion
> factor" for each "row". I've included my "work to date" below, but
I'm
> not very close to getting this right.
>
>
Without looking too closely, would it be possible to submit this
question to R-help again with a small example of what it is you are
looking for? For instance, construct a small, simulated data.frame
object like your 'real' one, and then show what you mean by "expansion
factors" (i.e, the output you'd like based on your simulated
data.frame). This usually helps get a quick, easy answer.
Hope that helps!
______________________________________________
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
and provide commented, minimal, self-contained, reproducible code.
