Err_f; convergency achieved

Dynare help (legacy posts)
1

Hi,
I’m trying to run a deterministic simulation of an OLG model.
Blanchard-Khan conditions are verified but I get the following message from Dynare :

"variable Dif period 1
1- err = 0.35502
err_f = 0.51026
Time of this iteration : 0.266
variable gr period 2
2- err = 0.033397
err_f = 0.079035
Time of this iteration : 0.235
variable gr period 2
3- err = 0.00076967
err_f = 0.0046939
Time of this iteration : 0.235
variable gr period 2
4- err = 2.5182e-006
err_f = 1.4305e-005
Time of this iteration : 0.235

Total time of simulation	: 1.062

Convergence achieved."

What does it means?
In particular:
why does Dynare report the name of some variables?
what is err_f and why do I get “convergence achieved” and not “convergency obtained”?
is there something wrong with the convergence of the model?

Thank you very much,
Best,

Alessandro

2

This is some debugging output that has now disappeared from the current versions. If I remember correctly, the variable name and period was indicating which variable in which period was most changed by the current Newton iteration.

Best

Michel

3

Dear Michel,
Thank you very much for your reply.
I’m glad to know that the messages I get are not related to a convergency problem of the code I wrote.

However I’m using Dynare version 3.064, that should be the last one of the “Dynare3 series”.

Has this kind of output (which only appears when I use a variable with lead 2) been changed in version 4 ?

Best

Alessandro

4

You are right. This form of output is still used in both versions when there are lags on more than one period.

I will remove it one of these days as it is more confusing than anything else.

Thanks for reporting it

Best

Michel