Forrtl: error(72): floating overflow?
From: "WANG,YANING" yaning@ufl.edu
Subject: [NMusers] Forrtl: error(72): floating overflow?
Date: 10/22/2003 1:03 PM
Dear users:
Could anyone suggest me some possible reasons for the following error to occur?
forrtl: error (72): floating overflow
Image PC Routine Line Source
ANIH2.exe 0040163D Unknown Unknown Unknown
ANIH2.exe 0040706C Unknown Unknown Unknown
ANIH2.exe 004727E3 Unknown Unknown Unknown
ANIH2.exe 00474DC2 Unknown Unknown Unknown
ANIH2.exe 0043BA46 Unknown Unknown Unknown
ANIH2.exe 0045920F Unknown Unknown Unknown
ANIH2.exe 004640F9 Unknown Unknown Unknown
ANIH2.exe 0042A390 Unknown Unknown Unknown
ANIH2.exe 0042090E Unknown Unknown Unknown
ANIH2.exe 004CBB09 Unknown Unknown Unknown
ANIH2.exe 004BDF24 Unknown Unknown Unknown
KERNEL32.dll 7C4E87F5 Unknown Unknown Unknown
With the exact same model and data, I got very good estimates (see below) for all the
parameters and SEs if FO is used. Once I switched to FOCE with interaction, this error
will show up and stop the fitting. Even after I put IF (ABS(ETA(1)).GT.10) EXIT 1 100
IF (ABS(ETA(2)).GT.10) EXIT 1 200
IF (ABS(ETA(3)).GT.10) EXIT 1 300
IF (ABS(ETA(4)).GT.10) EXIT 1 400
in my $PK block, the error messages still persist.
FO results:
THETA: CL-L/h V1-L Q-L/h V2-L
ETA: CL[P] V1[P] Q[P] V2[P]
ERR: [P] [A]
ANIH1.lst -5494.555 eval=320 sig=+3.4 sub=95 obs=2536 CCIL=NNNN NV1.
1 PIV1.1
THETA = 0.884 18.3 7.32 15.8
ETASD = 0.186011 0.279106 0.392428 0.185742
ERRSD = 0.104403 0.0519615
THETA:se% = 2.3 3.7 9.2 3.0
OMEGA:se% = 13.9 20.5 22.7 33.0
SIGMA:se% = 9.3 23.9
MINIMIZATION SUCCESSFUL
user 0:28.75 real 0:28.75 tcl 0:0.78
Thanks for any input.
Yaning