Quantcast
Channel: StackExchange Replication Questions
Viewing all articles
Browse latest Browse all 17268

Bootstrapping makeshift 2 stage cluster sampling code in R?

$
0
0

I have written code for a complex 2 staged sample in R, I want to compare mean estimates of the results from this code to other methods however, it is not a function or a dataset, simply code that creates the results. I have pasted the 2nd stage part my code which is being produced for multiple sample sizes, I want to complete the simulation w replication. I have attached code. Thanks for your help.

library(devtools) 
 source_gist(6424112)
en_min<-15
en_max<-100
SSS   <- matrix(c(rep(0,2*(en_max-en_min+1))),2,(en_max-en_min+1))
ps_min=2
psmax=13
 for (ps in ps_min:psmax) {
nabida<-merge(hoto,catte,by="Invoice..")
 ratatabin<-as.data.frame(table(pframe$Invoice..))
 shestamar<-as.data.frame(table(nabida$Invoice..))
colnames(shestamar)[1]<-c("Invoice..") 
rotodoto<-merge(nabida,shestamar,by="Invoice..")
beg2ndb<-stratified(rotodoto[order(rotodoto$Invoice..),], group = "Invoice..", size=ps)
lutut<-as.data.frame(table(beg2ndb$Invoice..))
colnames(lutut)[1]<-c("Invoice..")
colnames(shestamar)[1]<-c("Invoice..") 
 wytiq<-merge(beg2ndb,lutut,by="Invoice..")
 wytiq$secondselection<-wytiq$Freq.y/wytiq$Freq.x
 wytiq
 wytiq$finalweight<-(1/wytiq$secondselection)*wytiq$samwt
 wytiq

 en<-nrow(wytiq)
 ssuest<-sum(wytiq$finalweight*wytiq$COMMISSION..Profit.Ratio.)
 Tcritssu=qt(0.95, en-1)
 twatiri<-nrow(pframe)
lestamasia<-((twatiri-en)/(twatiri-37))**(((twatiri-37)/((twatiri-en)*((en/37)**2))))
ssuest-Tcritssu*sqrt(lestamasia*variance_ssupr)
hetadan<-ssuest-Tcritpsu*sqrt(lestamasia*variance_psupr)
SSS[1,en-(en_min-1)]=en
SSS[2,en-(en_min-1)]=ssuest-Tcritpsu*sqrt(lestamasia*variance_ssupr)

} 

Viewing all articles
Browse latest Browse all 17268

Trending Articles



<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>