Vithor Rosa Franco

Seven years later I am here to tell whoever may be interested in learning that the solution is rather simple: model { for ( i in 1:Nt ) { y[i] ~ dbern( pCorr[i] ) pCorr[i] <- ilogit( A[sID[i]] - D[iID[i]] ) } for ( i in 1:Ns ) { A[i] ~ dnorm( muA , sigmaA ) } for ( i in 1:Ni ) { D[i] ~ dnorm( muD , sigmaD ) } muA ~ dnorm(0,.1^2) sigmaA ~ dgamma(.01,.01) ProdAD ~ dgamma(.01,.01)T(C,) # Adjust the priors accordingly! muD ~ dnorm(0,.1^2) sigmaD <- ProdAD / sigmaA }

I need to fit a model where: SigmaA * SigmaD >= C It should be read as "the standard deviation of A, times the standard deviation of D, must be greater than or equal to a Constant". This rule have to be implemented in the following model model { for ( i in 1:Nt ) { y[i] ~ dbern( pCorr[i] ) pCorr[i] <- ilogit( A[sID[i]] - D[iID[i]] ) } for ( i in 1:Ns ) { A[i] ~ dnorm( muA , sigmaA ) } for ( i in 1:Ni ) { D[i] ~ dnorm( muD , sigmaD ) } muA ~ dnorm(0,.1^2) sigmaA ~ dgamma(.01,.01) muD ~ dnorm(0,.1^2)...

.

Martyn, Thank you for the prompt reply! I took all the seds out. Also, I deleted a part of the code that mentioned a 'tmp1' (which depended on a sed operation to exist). I spent the last month trying to run the code, making other little modifications. I was not sucessful, though. The error now is: MCMC for Stochastic Production Frontier Models Tue Jun 20 16:48:27 2017 Production Function: Cobb-Douglas Inefficiency term distribution: exp Iterations = 50001:100000 Thinning interval = 5 Sample size...

Martyn, Thank you for the prompt reply! I took all the seds out. Also, I deleted a part of the code that mentioned a 'tmp1' (which depended on a sed operation to exist). I spent the last month trying to run the code, making other little modifications. I was not sucessful, though. The error now is: MCMC for Stochastic Production Frontier Models Tue Jun 20 16:48:27 2017 Production Function: Cobb-Douglas Inefficiency term distribution: exp Iterations = 50001:100000 Thinning interval = 5 Sample size...

Martyn, Thank you for the prompt reply! I took all the seds out. Also, I deleted a part of the code that mentioned a 'tmp1' (which depended on a sed operation to exist). I spent the last month trying to run the code, making other little modifications. I was not sucessful, though. The error now is: MCMC for Stochastic Production Frontier Models Tue Jun 20 16:48:27 2017 Production Function: Cobb-Douglas Inefficiency term distribution: exp Iterations = 50001:100000 Thinning interval = 5 Sample size...

Thank you Martyn! I made your modifications, but still nothing. I have no idea what a sed is; never had a problem with that. I tried to find a solution online, but I don't know if it is R or JAGS related... I tried running the code on both Windows 10 and Windows 8.1 with JAGS 4.2.0 and R 3.4.0. Could you help me with that? What information do you need so you can help me with a north? Thanks again!