Dynare forums

Hi everybody,

I am solving a DSGE model using Bayesian Estimation with Dynare 4.3.2. Since the model is log-linearized, the steady state values are shown in the model equations as parameters.
Besides, to solve one of my parameter "eps", I need to use a Fsolve function. THerefore, I write a steady state file.
But when I run my mod, it always report error as follows:

Error: File: li_steadystate.m Line: 53 Column: 1
"eps" previously appeared to be used as a function or command, conflicting with its use here as the
name of a variable.
A possible cause of this error is that you forgot to initialize the
variable, or you have initialized it implicitly using load or eval.


"eps" is a parameter in my .mod file, and by using the steady state conditions, I have solved it as in the code.

Could anyone tell me how to settle the problem? THanks a lot!

Try giving it a different name. eps is a builtin matlab-function. This yields a naming conflict.

jpfeifer wrote:Try giving it a different name. eps is a builtin matlab-function. This yields a naming conflict.

Thanks for your reply! It works now.

However, some steady state values solved are complex numbers, and the estimation could not continue. THe error shows as follows:

Error in computing likelihood for initial parameter values
??? Error using ==> print_info at 65
Some updated params are complex

Could you please tell me what should I do now? Does it comes from my incorrect estimated parameters initialization? Are the estimated parameters necessarily be initialized?

I have posted my updated files.

THanks!

If you do not specify initial values for the parameters, Dynare takes the prior mean. Try providing explicit starting values.