From: "Luciane Velasque"
Subject: [NMusers] Ln(DV)
Date: Thu, 13 Mar 2003 10:45:30 -0300
I am trying to work with Ln(DV) and the ERROR record is :
$ERROR
IPRED=F
IF(IPRED.EQ.0) IPRED=EXP(-9)
Y=LOG(IPRED)+ERR(1)
ESTIMATION POSTHOC
The NONMEM returns me
0MINIMIZATION SUCCESSFUL
NO. OF FUNCTION EVALUATIONS USED: 4621
NO. OF SIG. DIGITS IN FINAL EST.: 5.1
0R MATRIX ALGORITHMICALLY NON-POSITIVE-SEMIDEFINITE
BUT NONSINGULAR
0COVARIANCE STEP ABORTED
0PRED EXIT CODE = 1 0INDIVIDUAL NO. 1 ID= .10000000E+01 (WITHIN-INDIVIDUAL) DATA REC NO. 1
THETA=
3.32E-02 2.22E-03 2.51E-05 9.00E-06 1.18E-01
OCCURS DURING SEARCH FOR ETA AT A NONZERO VALUE OF ETA
ERROR IN TRANS4 ROUTINE: Q IS NEGATIVE
0PROGRAM TERMINATED BY FNLETA
MESSAGE ISSUED FROM TABLE OR SCATTERPLOT STEP
What shall I do ?
Ln(DV)
From:"Bachman, William"
Subject: RE: [NMusers] Ln(DV)
Date:Thu, 13 Mar 2003 14:15:39 -050
Have you constrained Q to be positive?
example code for constraining PK parameters to be positive:
$SUBROUTINES ADVAN4 TRANS4
$PK
TVCL=THETA(1)
CL=TVCL*EXP(ETA(1))
TVV2=THETA(2)
V2=TVV2*EXP(ETA(2))
TVQ=THETA(3)
Q=TVQ*EXP(ETA(3))
TVv3=THETA(4)
V3=TVv3*EXP(ETA(4))
TVKA=THETA(5)
KA=TVKA*EXP(ETA(5))
S2=V2
;TLAG1=THETA(6)
;ALAG1=TLAG1*EXP(ETA(6))
$ERROR
IPRED=F
IF(IPRED.EQ.0) IPRED=EXP(-9)
Y=LOG(IPRED)+ERR(1)
$THETA
(0, 4.6) ;[CL]
(0, 6.5) ;[V2]
(0, 3.9) ;[Q]
(0, 3.0) ;[V3]
(0, 0.25);[KA]
;(0,0.2) ;[ALAG1]
From:Nick Holford [mailto:n.holford@auckland.ac.nz]
Subject:
Date:Thursday, MarchRe: [NMusers] Ln(DV 13, 2003 2:57 PM
Bill,
> example code for constraining PK parameters to be positive:
Hair splitting but...
This constrains parameters to be non-negative. Pop value for Q could become
0 couldn't it? Zero is not a positive number.
Nick
_______________________________________________________
From: "Venkatesh Atul Bhattaram"
Subject:Re: [NMusers] Ln(DV)
Date: Thu, 13 Mar 2003 10:24:28 -0500
Hello Lucaine
Could you share more information on the code and the data structure you have?
Venkatesh Atul Bhattaram
Post-doctoral Fellow
University of Florida
Gainesville-32610
From: Ludden, Thomas
Subject:RE: [NMusers] Ln(DV)
Date:Friday, March 14, 2003 8:41 AM
Bill,
The message is coming from a POSTHOC estimate of eta. Constraining theta's
does not constrain etas.
Tom