rpietro
commented
12 years ago would add the following after the recoding to avoid getting errors if your recoding is applied to rows you no longer have in the data after sampling:
nsqip.sample <- nsqip.data[sample(1:nrow(nsqip.data), 1000, replace=FALSE),] #replace 1000 with the number of observations you might want to keep, the smaller the faster, the greater the less likely you are to get a conversion error write.csv(nsqip.sample, file = "sample.csv") #run this if you want to save the sample. notice that it will save it to your working directory. to know where it is use getwd(). to set it see ?setwd
btw, you comment all those install.package lines, it takes an eternity to run and they are unnecessary
On Fri, Jun 29, 2012 at 9:38 AM, mworni < reply@reply.github.com
wrote:
Ricardo - could you please create a subset of the dataset to be able to play with the multiple imputation?
I guess that we will have to run the final imputation on your ferrari computer - mine won't be able to handle this I guess...
Reply to this email directly or view it on GitHub: https://github.com/rpietro/NSQIPageComplications/issues/23
Ricardo - could you please create a subset of the dataset to be able to play with the multiple imputation?
I guess that we will have to run the final imputation on your ferrari computer - mine won't be able to handle this I guess...