The analytic_derivation and loglinear options are not compat
Posted:
Mon Jul 25, 2016 8:26 am
by shuixing
dear professor Pfeifer
I build a stationary model and write the first order condition into dynare without loglinearize the variables and use a loglinear option in estimation set, and get a warning "The analytic_derivation and loglinear options are not compatible", it seems that there is a mistake using loglinear option, I am not clear where is wrong,what might be the problem, looking forward to hearing your reply, thanks.
best
Re: The analytic_derivation and loglinear options are not co
Posted:
Mon Jul 25, 2016 8:39 am
by jpfeifer
As it says, you cannot use the
- Code: Select all
loglinear
option together with the
- Code: Select all
analytic_derivation
option. If you have not enabled the latter on purpose, there must be something in your mod-file that triggers
- Code: Select all
analytic_derivation
before the estimation command. Please try adding
- Code: Select all
analytic_derivation=0
to the estimation command
Re: The analytic_derivation and loglinear options are not co
Posted:
Tue Jul 26, 2016 8:39 am
by shuixing
jpfeifer wrote:As it says, you cannot use the
- Code: Select all
loglinear
option together with the
- Code: Select all
analytic_derivation
option. If you have not enabled the latter on purpose, there must be something in your mod-file that triggers
- Code: Select all
analytic_derivation
before the estimation command. Please try adding
- Code: Select all
analytic_derivation=0
to the estimation command
dear professor peifer
I add analytic_derivation=0 to the estimation command, but it seems that Dynare does not accept the expression "analytic_derivation=0" in estimation (options), because it is reported that ',' is expected rather than '='.
Re: The analytic_derivation and loglinear options are not co
Posted:
Tue Jul 26, 2016 8:41 am
by jpfeifer
I see. Then add
- Code: Select all
options_.analytic_derivation=0;
immediately before the estimation command.
Re: The analytic_derivation and loglinear options are not co
Posted:
Thu Jul 28, 2016 3:22 am
by shuixing
thanks professor Peifer, it works.