[ADMB Users] .cov file

[Mark Maunder]

Richard,

Correlation( Sa, Sb ) = Covariance( Sa, Sb ) / (σa * σa). 

So you can check if the .cov file using the data from the .cor file.

Mark

 

Hi Mark,

 

Thank you again for your help. I have tried using different seeds to get MCMC chains and it worked for me. 

 

Now I have an enquiry on the .cov file. After an mpd run, admb generates file admodel.cov, which I guess contains the covariance matrix in binary. But I am not sure about it for two reasons: 1. the numbers look a little small to me; 2. in the examples of ADMB's user manual, there are functions to calculate covariance matrix. I would be very grateful if you could explain wether the contents of the .cov file are covariances and whether they are transformed or not. 

 

Thanks,

Richard

>>> Richard Bian 27/07/2010 11:08 a.m. >>>

Mark,

 

Thank you very much for your suggestion. I will give it a go and see what I get.

 

Regards,

Richard

>>> "Mark Maunder" <mmaunder at iattc.org> 27/07/2010 10:55 a.m. >>>

Richard,

You can use the -mcseed N command line option to generate different mcmc chains. I think it will start at the same place (the mode of the joint posterior) but will take a different path based on a different set of random numbers.

I hope this helps,

Mark

 

 

 

Mark Maunder 

Head of the Stock Assessment Program

Inter-American  Tropical Tuna Commission
8604 La Jolla Shores Drive
La Jolla, CA, 92037-1508, USA
  
Tel: (858) 546-7027
Fax: (858) 546-7133
mmaunder at iattc.org
http://www.fisheriesstockassessment.com/TikiWiki/tiki-index.php?page=Mark+Maunder

 

Visit the AD Model Builder project at
 http://admb-project.org/ <http://admb-project.org/> 

 

See the following website for information on fisheries stock assessment

http://www.fisheriesstockassessment.com/

 

From: Richard Bian [mailto:r.bian at niwa.co.nz] 
Sent: Monday, July 26, 2010 3:05 PM
To: Mark Maunder
Subject: inquiry on MCMC runs in ADMB

 

Dear Mark,

 

To evaluate MCMC chains for a model I built using ADMB, I need to set start points (eg, add random displacements to MPD) to an MCMC chain. Could you please suggest what I should do to set specific start points for an MCMC run. 

 

Thank you in advance!

 

Richard

 

 

Dr. Richard Bian
Scientist - Fisheries Modelling
National Institute of Water & Atmospheric Research
41 Market Place
Auckland Central 1010, New Zealand
Tel: +64 9 375 2054  Fax: +64 9 375 2051

 

 

NIWA is the trading name of the National Institute of Water & Atmospheric Research Ltd. 

 

NIWA is the trading name of the National Institute of Water & Atmospheric Research Ltd. 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.admb-project.org/pipermail/users/attachments/20100812/c53017ea/attachment.html>