Selecting random data in SQL and R

I have this SQL-Server code: ABS(checksum(newid())%14

that generates normally distributed random numbers between 0 and 13.

More specifically, I have a column in my SQL table that has these random numbers. Doing some analysis in R, I have now imported this SQL table into R. I would like to create a test and training set on 80%-20% of this data based on this column. Do you know how I would do that? Would I need to subset where random number >=7, to create the training set?

Thanks.