Duplicate IDs in phenotype input crash step 2

[DDochtermann]

Took some digging, as the step 2 will fail and core dump with a floating point exception and not output a useful backtrace.

Step 1 will complete and output the modglmm Rdata, but using that object will crash step 2. Step 1 recognizes duplicate IDs as it writes the output log, but does not remove them or limit to unique instances. Recursively removing the duplicates from the nested list R object makes the step 1 output usable again.

Would recommend error-ing out step 1 or dropping duplicates to avoid the nondescript step 2 crashes.