RE: different COMRES statements

-----Original Message----- From: Alison Boeckmann [mailto:[EMAIL PROTECTED] Sent: Tuesday, October 16, 2007 4:22 PM To: Andrew Hooker; 'Frederik Gorter de Vries'; [email protected] Subject: RE: [NMusers] different COMRES statements Frederick: I don't understand the references to $PRIOR. This is not a feature of NONMEM VI 1.0 or 1.2, although it will be of 2.0. Perhaps you are referring to the new feature introduced with NONMEM VI: user-supplied PRIOR subroutine and utility programs NWPRI and TNPRI. How does the use of COMRES interact with the use of these subroutines? COMRES and the COM(i) variables are used to provide values for NONMEM to display in tables, and can also be used to preserve values of the same variable between different passes through the data (e.g., between passes when etas are 0 vs. conditional.) Are you perhaps attempting to put values of CNT (the contribution to the objective function computed by NWPRI or TNPRI) in a table? Or are you using some other feature of NONMEM that involves use of COM(i), and $PRIOR was a typing error? Andrew, I agree with everything you wrote, except that I can't understand what it has to do with "$PRIOR functionality in NONMEM." Perhaps you mean "PRIOR functionality", and the difficulty is that, when all of OMEGA is written, some of the OMEGA values were fixed values supplied in the control stream for computation of the priors, and are of no interest in the output table. I can see the difficulty, but cannot understand its relationship to the COMRES/COM(i) feature. On Mon, 15 Oct 2007 15:50:17 +0200, "Andrew Hooker" <[EMAIL PROTECTED]> said: > Hi Frederik, > > > > It is possible to compute the CWRES using the $PRIOR functionality in > NONMEM. However, it's a little tricky. The problem is this code: > > > > $INFN > > IF (ICALL.EQ.3) THEN > > OPEN(50,FILE=CWTAB1.EST) > > WRITE(50,*) 'ETAS' > > DO WHILE(DATA) > > > IF (NEWIND.LE.1) WRITE (50,*) ETA > > > ENDDO > > > WRITE(50,*) 'THETAS' > > WRITE(50,*) THETA > > WRITE(50,*) 'OMEGAS' > > WRITE(50,*) OMEGA(BLOCK) > > WRITE(50,*) 'SIGMAS' > > WRITE(50,*) SIGMA(BLOCK) > > ENDIF > > > > Which writes out all the thetas and omegas and sigmas including the > prior stuff (which we don't need for the CWRES calculation). What I > do is just remove, by hand, the extra values in the "cwtab1.est" file > to have only the values for the etas, thetas, omegas and sigmas for > the data and not the prior (only the "first" thetas, omegas and > sigmas). After that, calculating the CWRES using XPOSE 4 should not > be a problem. > > > > For the COMRES values: If you need x COMRES values for the CWRES > calculation and y COMRES values for some other portion of your NONMEM > model then you just need to use the statement > > > > $ABB COMRES= z > > > > Where z = x+y. > > > > Hope this helps! > > > > -Andy > > > > > > Andrew Hooker, Ph.D. > > Assistant Professor of Pharmacometrics > > Div. of Pharmacokinetics and Drug Therapy > > Dept. of Pharmaceutical Biosciences > > Uppsala University > > Box 591 > > 751 24 Uppsala > > Sweden > > Tel: +46 18 471 4355 > > www.farmbio.uu.se/research.php?avd=5 > > > > _____ > > From: [EMAIL PROTECTED] [mailto:owner- > [EMAIL PROTECTED] On Behalf Of Frederik Gorter de Vries Sent: > Monday, October 15, 2007 1:40 PM To: [email protected] Subject: > [NMusers] different COMRES statements > > > > Dear NMuser, > > > > I have a problem with implementing the CWRES in a model that's using > $PRIOR. Both need a $ABB COMRES= x statement, but with different > numbers of x. Which number should I take here? NONMEM seems to have no > problem when using the one from CWRES, but the CWRES cannot be > calculated anymore in R using Xpose. The following message appears: > > > > > cwres <- compute.cwres(1, printToOutFile = FALSE,onlyNonZero = > > FALSE) > > Error in `[.data.frame`(x, r, vars, drop = drop) : > > undefined columns selected > > > > When I use the COMRES number from the $PRIOR NMTRAN gives an error > regarding the table with the wrong number of COM ARRAY . > > > > I hope someone have seen this problem before and can tell me more > about the COMRES statement. > > > > Thank you in advance and best regards, > > > > Frederik Gorter de Vries > > > > > > _______________________________________ > > Frederik Gorter de Vries > > LACDR / Pharmacology > > Gorleaus Laboratories > > Einsteinweg 55 2333 CC Leiden > > The Netherlands > > Tel : ++31 71 527 6142 > > Email: [EMAIL PROTECTED] > > > > > -- Alison Boeckmann [EMAIL PROTECTED]