I have a file in which I have 1 trait (DBH) for two different sets of
hybrids - I came up with some code which I thought would create two
variates (DBHPOOC & DBHPTEC) for the two sorts of hybrids (PTEC & POOC) to
eventually run a bivariate model
Title Line: PCH x POOC & PTEC QFRI Trial 690 Site WS 5 yo
HYBRID 2 !A # Two levels, alphabetic Coding: PTEC & POOC
MUM 12
DAD 12
... #
and a few others I've dropped
WF HT DBH VOL LN ST D1 D2 D3 D4 D5 D6 D7 # Traits
NOTRUNT !V21=V10 !>$A !D 0 # Drops records with
DBH < input DBH limit
DBHPOOC !V22=V1 !==1 !*V10 !M 0 # Create DBH for POOC
i.e. HYBRID level 1
DBHPTEC !V23=V1 !==2 !*V10 !M 0 # Create DBH for PTEC
Q690WS.DAT !skip 1 # Data file name, skip first record
# fit model
DBHPOOC ~ mu REP !r INBLK MUM DAD MUM.DAD
0
Now this seems to work fine for DBHPOOC but when I try to run for DBHPTEC
it thinks there are no records. If I dont make zeroes missing then the
number of zeroes for DBHPTEC & DBHPOOC indicates that it seems to be
picking up the right records for each hybrid type. Even wierder, if I
reverse the order of creation of DBHPOOC & DBHPTEC and run for DBHPTEC,
then it runs but it is clearly using the DBHPOOC data!