Bug in imcforecast.m?

Hi,

I am running the snapshot version of Dynare version 4.

I am trying to do conditional forecasting, as described here: dynare.org/DynareWiki/ConditionalForecast

I obtained the following error message:

[code]??? Input argument “missing_value” is undefined.

Error in ==> DsgeSmoother at 259
if missing_value

Error in ==> imcforecast at 40
[atT,innov,measurement_error,filtered_state_vector,ys,trend_coeff] =
DsgeSmoother(xparam,gend,data);[/code]

Replacing line 40 in imcforecast.m with the following seems to solve the problem:

[atT,innov,measurement_error,filtered_state_vector,ys,trend_coeff] = DsgeSmoother(xparam,gend,data,],0);

Is this a bug? Is my “mechanical solution” correct?

Thanks,

Riccardo

Hi Riccardo, your solution is correct. I did forget to change this file when i introduced the possibility to estimate models with missing observations. I will fix this as soon as possible. Thanks for your bug report.

Best,
Stéphane.