From:"Kyun-Seop Bae"
Subject:[NMusers] Error Message: Initial Values for ATOL ?
Date:Sun, February 10, 2002 6:22 am
Hi, everyone,
While I was modeling with ADVAN 9, (including 1 equilibrium comparment)
I got an error message of
' OCCURS DURING SEARCH FOR ETA AT INITIAL VALUE, ETA=0
ERROR OCURRED WHILE ATTEMPTING TO OBTAIN INITIAL VALUES FOR ATOL'.
I could not find help on 'ATOL' in nmhelp.
Could anyone suggest me what is wrong with my following control code or data file ?
Sincerely,
Kyun-Seop Bae MD
Clinical Trial Center
Seoul National University Hospital
28 Yongon-dong Chongno-gu
Seoul, 110-744, Korea
Tel: +82-2-760-1666
Email: ksbae@snu.ac.kr
===========================
// OUTPUT FILE WITH ERROR MESSAGES //
0PRED EXIT CODE = 1
0INDIVIDUAL NO. 1 ID=0.10000000E+01 (WITHIN-INDIVIDUAL) DATA REC NO. 5
THETA= 5.00E-01 2.00E+00 1.00E-02
OCCURS DURING SEARCH FOR ETA AT INITIAL VALUE, ETA=0
ERROR IN LSODI1: CODE 205
ERROR OCURRED WHILE ATTEMPTING TO OBTAIN INITIAL VALUES FOR ATOL
0PROGRAM TERMINATED BY OBJ
MESSAGE ISSUED FROM ESTIMATION STEP
AT INITIAL OBJ. FUNCTION EVALUATION
1THERE ARE ERROR MESSAGES IN FILE PRDERR
Stop - Program terminated.
===========================
// CONTROL FILE //
$PROBLEM MODEL XXX
$INPUT ID TIME AMT DV CMT
$DATA A9_2.CSV IGNORE=#
$SUBROUTINE ADVAN9 TOL=10
$MODEL
NCOMP = 5 ;
COMP = (WATER) ; A(1), W
COMP = (ACH) ; A(2), X
COMP = (AC) ; A(3), YY
COMP = (HCL) ; A(4), Z
COMP = (H, DEFOBS, EQU) ; A(5), H
$PK
PKA = 5.5;
KA = 10 ** (-1 * PKA); ; Cause of error is too frequent use of
exponentials ?
VMAX = THETA(1)*EXP(ETA(1)) ; Error does not change with additive eta
KM = THETA(2);
KGE = THETA(3);
$AESINIT
INIT = 0; 0=APPROXIMATE, 1=EXACT
A(5) = 0.2;
$AES
E(5) = A(5)*(A(3)-A(4)+A(5))/(A(2)+A(4)-A(5)) - KA * A(1) ; Is this too complex?
$DES
PH = -1 * LOG (A(5)/A(1))
DADT(1) = -1 * KGE * A(1)
DADT(2) = -1 * KGE * A(2)
DADT(3) = -1 * KGE * A(3)
DADT(4) = VMAX * PH / (KM + PH) - KGE * A(4)
$ERROR
Y = F * EXP(EPS(1))
$THETA
(0, 0.5) ;THETA(2), VMAX - Is my initial guesses not adequate?
(0, 2) ;THETA(3), KM
(0, 0.01) ;THETA(4), KGE
$OMEGA
0.2
$SIGMA
0.02
$ESTIMATION METHOD=1 ; No matter with estimation method. the same error
message
===============================
// DATA FILE, CSV FORMAT IN ORIGINAL//
#ID TIME AMT DV CMT
1 0 0.9 . 1
1 0 0.9 . 2
1 0 0.11 . 3
1 0 0.4 . 4
1 1 . 5.37E-06 4
1 2 . 1.13E-05 4
1 3 . 1.43E-05 4
1 4 . 3.87E-06 4
1 5 . 4.92E-06 4
1 6 . 6.50E-06 4
1 7 . 9.16E-06 4
1 8 . 1.45E-05 4
1 9 . 3.08E-05 4
1 10 . 0.003155421 4
1 11 . 0.286965292 4
1 12 . 0.123910793 4
1 13 . 0.2 4
2 0 1 . 1
2 0 1 . 2
2 0 0.1 . 3
2 0 0.5 . 4
2 1 . 0.00001 4
2 2 . 0.00001 4
2 3 . 1.58E-05 4
2 4 . 3.16E-06 4
2 5 . 5.01E-06 4
2 6 . 6.31E-06 4
2 7 . 0.00001 4
2 8 . 1.58E-05 4
2 9 . 3.16E-05 4
2 10 . 0.003162278 4
2 11 . 0.316227766 4
2 12 . 0.1 4
2 13 . 0.199526231 4
Error Message: Initial Values for ATOL ?
From:"Sam Liao"
Subject:RE: [NMusers] Error Message: Initial Values for ATOL ?
Date:Sun, February 10, 2002 3:51 pm
Hi Kyun-Seop:
There may be some problem in your model specification. When I tried your
control file without the estimation step, I got
the following error msg. This indicated the problem is likely in your model
specification.
==============================================
WARNINGS AND ERRORS (IF ANY) FOR PROBLEM 1
(WARNING 2) NM-TRAN INFERS THAT THE DATA ARE POPULATION.
1 file(s) copied.
Starting nonmem execution ...
forrtl: error (65): floating invalid
===============================================
Best regards,
Sam Liao, Ph.D.
PharMax Research
270 Kerry Lane,
Blue Bell, PA 19422
phone: 215-6541151
efax: 1-720-2946783
From:"Kyun-Seop Bae"
Subject:[NMusers] Error: Initial Values for ATOL / LSODI1 CODE 205 ?
Date:Sun, February 10, 2002 8:30 pm
Dear Juan Jose and Sam,
Thank you very much for your immediate help.
CMT=4 was typographical error, and with the correction of this, the problem still
remains.
TOL=5 or 6 was also failed, I increased it.
Fixing eta as 0 also leads the same error.
For Sam's advice, I don't have your error message without estimation step.
But, similar to the previous one : 'ERROR IN LSODI1: CODE 205'
Maybe this is more fundamental cause than previous mail.
I found the fortran code in the file of 'LSODI1.for', but it is not easy to figure out.
(Just I found that state of KFLAG variable determines the error code.)
Sincerely,
Kyun-Seop Bae MD
Clinical Trial Center
Seoul National University Hospital
28 Yongon-dong Chongno-gu
Seoul, 110-744, Korea
Tel: +82-2-760-1666
Email: ksbae@snu.ac.kr
===========================
// OUTPUT FILE WITH ERROR MESSAGES //
0PRED EXIT CODE = 1
0INDIVIDUAL NO. 1 ID=0.10000000E+01 (WITHIN-INDIVIDUAL) DATA REC NO. 5
THETA= 5.00E-01 2.00E+00 1.00E-02
OCCURS DURING SEARCH FOR ETA AT INITIAL VALUE, ETA=0
ERROR IN LSODI1: CODE 205
ERROR OCURRED WHILE ATTEMPTING TO OBTAIN INITIAL VALUES FOR ATOL
0PROGRAM TERMINATED BY OBJ
MESSAGE ISSUED FROM ESTIMATION STEP
AT INITIAL OBJ. FUNCTION EVALUATION
1THERE ARE ERROR MESSAGES IN FILE PRDERR
Stop - Program terminated.
===========================
// CONTROL FILE //
$PROBLEM MODEL XXX
$INPUT ID TIME AMT DV CMT
$DATA A9_2.CSV IGNORE=#
$SUBROUTINE ADVAN9 TOL=10 ; TOL=5 or 6 shows the same error
$MODEL
NCOMP = 5 ;
COMP = (WATER) ; A(1), W
COMP = (ACH) ; A(2), X
COMP = (AC) ; A(3), YY
COMP = (HCL) ; A(4), Z
COMP = (H, DEFOBS, EQU) ; A(5), H
$PK
PKA = 5.5;
KA = 10 ** (-1 * PKA); ; Cause of error is too frequent use of
exponentials ?
VMAX = THETA(1)*EXP(ETA(1)) ; Error does not change with additive eta
KM = THETA(2);
KGE = THETA(3);
$AESINIT
INIT = 0; 0=APPROXIMATE, 1=EXACT
A(5) = 0.2;
$AES
E(5) = A(5)*(A(3)-A(4)+A(5))/(A(2)+A(4)-A(5)) - KA * A(1) ; Is this too complex?
$DES
PH = -1 * LOG (A(5)/A(1))
DADT(1) = -1 * KGE * A(1)
DADT(2) = -1 * KGE * A(2)
DADT(3) = -1 * KGE * A(3)
DADT(4) = VMAX * PH / (KM + PH) - KGE * A(4)
$ERROR
Y = F * EXP(EPS(1))
$THETA
(0, 0.5) ;THETA(2), VMAX - Is my initial guesses not adequate?
(0, 2) ;THETA(3), KM
(0, 0.01) ;THETA(4), KGE
$OMEGA
0.2
$SIGMA
0.02
$ESTIMATION METHOD=1 ; No matter with estimation method. the same error
message
===============================
// DATA FILE, CSV FORMAT IN ORIGINAL// -> TYPO CORRECTED(CMT 4 -> 5)
#ID TIME AMT DV CMT
1 0 0.9 . 1
1 0 0.9 . 2
1 0 0.11 . 3
1 0 0.4 . 4
1 1 . 5.37E-06 5
1 2 . 1.13E-05 5
1 3 . 1.43E-05 5
1 4 . 3.87E-06 5
1 5 . 4.92E-06 5
1 6 . 6.50E-06 5
1 7 . 9.16E-06 5
1 8 . 1.45E-05 5
1 9 . 3.08E-05 5
1 10 . 0.003155421 5
1 11 . 0.286965292 5
1 12 . 0.123910793 5
1 13 . 0.2 5
2 0 1 . 1
2 0 1 . 2
2 0 0.1 . 3
2 0 0.5 . 4
2 1 . 0.00001 5
2 2 . 0.00001 5
2 3 . 1.58E-05 5
2 4 . 3.16E-06 5
2 5 . 5.01E-06 5
2 6 . 6.31E-06 5
2 7 . 0.00001 5
2 8 . 1.58E-05 5
2 9 . 3.16E-05 5
2 10 . 0.003162278 5
2 11 . 0.316227766 5
2 12 . 0.1 5
2 13 . 0.199526231 4
From:"Piotrovskij, Vladimir [PRDBE]"
Subject:RE: [NMusers] Error: Initial Values for ATOL / LSODI1 CODE 205 ?
Date:Mon, February 11, 2002 3:56 am
Title: RE: [NMusers] Error: Initial Values for ATOL / LSODI1 CODE 205 ?
Kyun-Seop,
ERROR IN LSODI1: CODE 205 or 204 means DES does not feel comfortable
with your model and/or (initial) parameter values.
I think your model is not fully correct as it returns negative A(5):
the first error I got when I ran your code "as is" was
run-time error M6201: MATH
- log: DOMAIN error
When I fixed PH = 7 in your $DES block and ran $SIMULATION ONLY
(OMEGA and SIGMA fixed to 0) I got the message
ERROR IN LSODI1: CODE 205. After playing with initial values
I found that with
$THETA
(0, 0.5) ;THETA(2), VMAX - Is my initial guesses not adequate?
(0, .2) ;THETA(3), KM
(0, 0.01) ;THETA(4), KGE
$SIMULATION can be done. Examining As revealed A(5)<0. It seems you have to check the model.
Best regards,
Vladimir
-----------------------------------------------------------------
Vladimir Piotrovsky, Ph.D.
Research Fellow, Advanced PK-PD Modeling & Simulation
Global Clinical Pharmacokinetics and Clinical Pharmacology (ext.
5463/151)
Johnson & Johnson Pharmaceutical Research & Development
Turnhoutseweg 30
B-2340 Beerse
Belgium
Fax: (+3214) 605834
Email: vpiotrov@prdbe.jnj.com
From: "Kyun-Seop Bae"
Subject:Re: [NMusers] Error: Initial Values for ATOL / LSODI1 CODE 205 ? - Yaho !
Date:Tue, February 12, 2002 1:39 am
RE: [NMusers] Error: Initial Values for ATOL / LSODI1 CODE 205 ?Dear All, and
especially Vladimir,
I found the clue from Vladimir's suggestion.
In $AES block,
E(5) can have too solutions, that is to say quadratic function with regard to A(5).
One is that A(5) > 0 , and the other is A(5) < 0.
So, if I replace the quadratic equation with the following always positive root,
then the error has gone.
=================
$AES
B = KA * A(1) + A(3) - A(4);
C = -1 * KA * A(1) * ( A(2) + A(4) );
D = B ** 2 - 4 * C;
E(5) = A(5) - 0.5 * ( -1 * B + D ** 0.5 );
=================
Thank all who considered my problem, and especially Vladimir.
And dear Dr. Kang, I'll be at Atlanta this march for ASCPT 2002.
Sincerely
Kyun-Seop Bae MD
Clinical Trial Center
Seoul National University Hospital
28 Yongon-dong Chongno-gu
Seoul, 110-744, Korea
Tel: +82-2-760-1666
Email: ksbae@snu.ac.kr