RE: Covariance matrix of parameter estimates (standard errors)
Hi Nick,Thanks a lot for your help. The replacement nmbs.bat works as you
described. I agree with your comment about the robustness of bootstrap
estimates in general, but my particular application requires the covariance
matrix for each run. On a related note, as part of development efforts for
NONMEM, it would really be helpful if any NONMEM estimates which are available
in the .lst file can be included in a $TABLE file. I realize there may be
clever ways to do this using abbreviated code but these are usually accessible
to more advanced users.
Kind regards,JO
> Date: Sun, 14 Jul 2013 07:16:14 +0200
> From: [email protected]
> To: [email protected]
> Subject: Re: [NMusers] Covariance matrix of parameter estimates (standard
> errors)
>
> JO,
>
> Sorry - I didn't understand your earlier request because I had not ever
> thought anybody would be interested in those values.IMHO there is almost
> no reason to collect this information because the bootstrap parameter
> estimates will give you a more reliable variance-covariance matrix of
> the estimates.
>
> I will send you separately a replacement nmbs.bat file for use with WFN.
> This collects any .cov file that is produced as a single file which will
> be in the results directory when nmgosim (or nmbs) finishes. You have to
> set the nmraw variable to n or y in order to get this behaviour.
>
> Its up to you to figure out how to extract and use the numbers :-)
>
> Best wishes
>
> Nick
>
Quoted reply history
> On 14/07/2013 3:32 a.m., NIYI ADEDOKUN wrote:
> >
> > Hi Nick,
> >
> > Thanks for your response. I realize the RSEs of the parameter
> > estimates are included in the .smy file but what I actually need is
> > the variance-covariance matrix of the estimates. For example, I am
> > thinking of a table containing the variance of theta(1), variance of
> > theta(2) as well as the covariance between theta(1) & theta(2). These
> > values are usually provided in the .lst file following an estimation
> > run but I would like to have them for all the simulation/estimation
> > runs with a successful covariance step. Since these estimates are in
> > the .lst file, there should be some way of outputting them to a table
> > for further analysis.
> >
> > Kind regards,
> >
> > JO
> >
> >
> > > Date: Sat, 13 Jul 2013 10:06:38 +0200
> > > From: [email protected]
> > > To: [email protected]
> > > Subject: Re: [NMusers] Covariance matrix of parameter estimates
> > (standard errors)
> > >
> > > JO,
> > >
> > > *IF* NONMEM is able to complete the covariance step then the relative
> > > standard errors are included in the WFN nmgosim output file.
> > >
> > > You will not find the RSE values if a) you did not include $COV in the
> > > control stream or b) NONMEM failed in the covariance step.
> > >
> > > You can test nmgosim with WFN by running:
> > >
> > > nmgosim theopd_gosim theopd_gest 1 5
> > >
> > > in the WFN run directory. You will find all the parameters and their
> > RSE
> > > values in run\theopd_goest.reg\theopd_goest.smy
> > >
> > > Usually there is little interest in calculating these asymptotic RSE
> > > values because the parametric bootstrap performed by nmgosim gives more
> > > reliable estimates of the RSE and confidence intervals of parameters.
> > >
> > > Nick
> > >
> > > On 12/07/2013 10:09 p.m., NIYI ADEDOKUN wrote:
> > > > Dear All,
> > > > Is there a simple way to include the standard errors of parameter
> > > > estimates (as found in the .cov file) in an output table following
> > > > sequential simulation and estimation in NONMEM? On a related note,
> > the
> > > > NMGOSIM command in Wings for NONMEM does not appear to include the
> > > > .cov file in its output - is there a workaround such that the
> > standard
> > > > errors of estimates (e.g. SETHET) can be included in the summary file?
> > > >
> > > > Kind regards,
> > > > JO
> > > >
> > > >
> > >
> > > --
> > > Nick Holford, Professor Clinical Pharmacology
> > > Dept Pharmacology & Clinical Pharmacology, Bldg 503 Room 302A
> > > University of Auckland,85 Park Rd,Private Bag 92019,Auckland,New Zealand
> > > office:+64(9)923-6730 mobile:NZ +64(21)46 23 53 FR +33(7)85 36 84 99
> > > email: [email protected]
> > > http://holford.fmhs.auckland.ac.nz/
> > >
> > > Holford NHG. Disease progression and neuroscience. Journal of
> > Pharmacokinetics and Pharmacodynamics. 2013;40:369-76
> > http://link.springer.com/article/10.1007/s10928-013-9316-2
> > > Holford N, Heo Y-A, Anderson B. A pharmacokinetic standard for
> > babies and adults. J Pharm Sci. 2013:
> > http://onlinelibrary.wiley.com/doi/10.1002/jps.23574/abstract
> > > Holford N. A time to event tutorial for pharmacometricians. CPT:PSP.
> > 2013;2: http://www.nature.com/psp/journal/v2/n5/full/psp201318a.html
> > > Holford NHG. Clinical pharmacology = disease progression + drug
> > action. British Journal of Clinical Pharmacology. 2013:
> > http://onlinelibrary.wiley.com/doi/10.1111/bcp.12170/abstract
> > >
> > >
>
> --
> Nick Holford, Professor Clinical Pharmacology
> Dept Pharmacology & Clinical Pharmacology, Bldg 503 Room 302A
> University of Auckland,85 Park Rd,Private Bag 92019,Auckland,New Zealand
> office:+64(9)923-6730 mobile:NZ +64(21)46 23 53 FR +33(7)85 36 84 99
> email: [email protected]
> http://holford.fmhs.auckland.ac.nz/
>
> Holford NHG. Disease progression and neuroscience. Journal of
> Pharmacokinetics and Pharmacodynamics. 2013;40:369-76
> http://link.springer.com/article/10.1007/s10928-013-9316-2
> Holford N, Heo Y-A, Anderson B. A pharmacokinetic standard for babies and
> adults. J Pharm Sci. 2013:
> http://onlinelibrary.wiley.com/doi/10.1002/jps.23574/abstract
> Holford N. A time to event tutorial for pharmacometricians. CPT:PSP. 2013;2:
> http://www.nature.com/psp/journal/v2/n5/full/psp201318a.html
> Holford NHG. Clinical pharmacology = disease progression + drug action.
> British Journal of Clinical Pharmacology. 2013:
> http://onlinelibrary.wiley.com/doi/10.1111/bcp.12170/abstract
>
>