Mark Knecht
2009-Jul-23 18:04 UTC
[R] howto create a list row-by-row as input to function call?
Hi,
I'm having trouble within my function CalcPos to get it to call
CalcHorz with values from each row. I *think* it's calling CalcHorz
with the final values of the inputs and not the values from each row.
How can I do this properly in R?
The values aa,bb,cc,dd are inputs. CalcPos first calculates V1 and
V2 vertically, and then I attempt to call CalcHorz to handle H1, H2 &
H3 using a list L1 comprised of df[aa:V2]. I wanted that list to be
the values on that row and disposed of each pass, but I think all
calls are probably using the final values. Finally I use the returned
values H1:H3 for more vertical calculations in V3.
How can I get list L1 to be the unique row values for each call to
CalcHorz? L1 would just be tossed when the call to CalcPos is
complete.
PLEASE - do not combine any of the inputs or return values. I need
all of them aa through V3 and want to keep them separate. Thanks!
Cheers,
Mark
CalcHorz = function (L1) {
# Intended to calculate 3 values for each row
H1 <- with(L1, V1*cc)
H2 <- with(L1, V2*dd)
H3 <- with(L1, H1 - H2)
return(as.list(H1, H2, H3))
}
CalcPos = function (df) {
#calculate vertical initially
df$V1 <- with(df, aa*bb)
df$V2 <- with(df, aa*cc-dd)
#At this point do 3 horizontal calculations
#Goal is this is done for each row using
#the values aa,bb,cc,dd,V1,V2 for that row
L1 = as.list(df[1:dim(df)[2]])
R1 = as.list(with(df, CalcHorz(L1)))
df$H1 <- R1[1]
df$H2 <- R1[2]
df$H3 <- R1[3]
#Switch back to vertical
df$V3 = with (df, cumsum(H3))
return(df)
}
DF <- data.frame(aa=1:12, bb=12:15, cc=1:3, dd=1:6)
DF
DF <- CalcPos(DF)
DF
Mark Knecht
2009-Jul-23 19:28 UTC
[R] howto create a list row-by-row as input to function call?
On Thu, Jul 23, 2009 at 11:04 AM, Mark Knecht<markknecht at gmail.com> wrote:> Hi, > ? I'm having trouble within my function CalcPos to get it to call > CalcHorz with values from each row. I *think* it's calling CalcHorz > with the final values of the inputs and not the values from each row. > How can I do this properly in R? > > ? The values aa,bb,cc,dd are inputs. CalcPos first calculates V1 and > V2 vertically, and then I attempt to call CalcHorz to handle H1, H2 & > H3 using a list L1 comprised of df[aa:V2]. I wanted that list to be > the values on that row and disposed of each pass, but I think all > calls are probably using the final values. Finally I use the returned > values H1:H3 for more vertical calculations in V3. > > ? How can I get list L1 to be the unique row values for each call to > CalcHorz? L1 would just be tossed when the call to CalcPos is > complete. > > ? PLEASE - do not combine any of the inputs or return values. I need > all of them aa through V3 and want to keep them separate. Thanks! > > Cheers, > Mark >Doing the CalcHorz as a data.frame solved the problem for me. Sorry for the noise. - Mark CalcHorz = function (L1) { # Intended to calculate 3 values for each row H1 <- with(L1, V1*cc) H2 <- with(L1, V2*dd) H3 <- with(L1, H1 - H2) return(data.frame(H1, H2, H3)) } CalcPos = function (df) { #calculate vertical initially df$V1 <- with(df, aa*bb) df$V2 <- with(df, aa*cc-dd) #At this point do horizontal calculations #Goal is this is done for each row using #the values aa,bb,cc,dd,V1,V2 for that row L1 = df[1:dim(df)[2]] R1 = CalcHorz(L1) df$H1 <- R1[1] df$H2 <- R1[2] df$H3 <- R1[3] #Switch back to vertical df$V3 = with (df, cumsum(H3)) return(df) } DF <- data.frame(aa=1:12, bb=12:15, cc=1:3, dd=1:6) DF DF <- CalcPos(DF) DF