[R] replace loops with matrix

Hey Sky heyskywalker at yahoo.com
Wed Aug 18 00:22:28 CEST 2010


Hey, R users

I am using numerical method for my research paper and the computation burden is 
very heavy. first I tried to do it with loops, example code as following, and it 
take hours to converge for only 200 obs. and my real data has 4000 obs. and the 
optimization command that I use is: 


optim(guess,myfunc1,data=mydata, method="BFGS",hessian=T))

then I tried matrix form computation, it takes only 1/10 of the time the loop 
method takes. it may still have room to improve it. at least, the following 
part looks ugly.
ccl[,m]<-lia[,1]*lia[,2]*lia[,3]*lia[,4]*lia[,5] 

any suggestion are appreciated.


The Loop code:
for(m in 1:ns){
 for(i in 1:nt){
    vbar2[,i]=a[1]+     eta[m]+acedu[,i]*a[2]+acwrk[,i]*a[3]
    vbar3[,i]=b[1]+b[2]*eta[m]+acedu[,i]*b[3]+acwrk[,i]*b[4]

    v8[,i]=1+exp(vbar2[,i])+exp(vbar3[,i])

      for(j in 1:n){
          if (edu[j,i]==1) lia[j,i]=1/v8[j,i]
          if (wrk[j,i]==1) lia[j,i]=exp(vbar2[j,i])/v8[j,i]
          if (home[j,i]==1) lia[j,i]=exp(vbar3[j,i])/v8[j,i]
     }
   ccl[,m]<-lia[,i]*ccl[,m]
 }
}

The Matrix code:
for(m in 1:ns){
    vbar2[,1:nt]=a[1]+     eta[m]+acedu[,1:nt]*a[2]+acwrk[,1:nt]*a[3]
    vbar3[,1:nt]=b[1]+b[2]*eta[m]+acedu[,1:nt]*b[3]+acwrk[,1:nt]*b[4]

    v8[,1:nt]=1+exp(vbar2[,1:nt])+exp(vbar3[,1:nt])

    lia[1:n,]<-ifelse(edu[1:n,]==1,1/v8[1:n,],
                           ifelse(wrk[1:n,]==1,exp(vbar2[1:n,])/v8[1:n,],
                                                       
exp(vbar3[1:n,])/v8[1:n,]))

   ccl[,m]<-lia[,1]*lia[,2]*lia[,3]*lia[,4]*lia[,5]
}

Nan
from Montreal





More information about the R-help mailing list