#125 Johansen misleading output with exogenous variables

[Sven S.]

the exogenous variables in a Johansen test setting aren't reported; and a warning should be printed that the critical values and p-values are in general only appropriate in the case without exogenous variables

[Allin Cottrell]

Please see if what is now in CVS is acceptable.

[Sven S.]

Sorry for the delay, Allin! I just tested with gretl 1.8.7, and for the case of unrestrited exo regressors it's fine. But I would suggest to print the same stuff for restricted exo regressors as well. So not closing this yet.

thanks,
sven

[Sven S.]

just an update, no change with restricted regressors in today's 1.8.7cvs.

Actually, a related suggestion for when this will be fixed: Maybe in addition to 'number of equations =' also print out the names of the endogenous variables; because with included restricted exogenous variables it's not 100% clear in the Johansen output which ones are the endogenous ones. (Of course I guess it's simply the last ones in the beta output, but still for clarification it may make sense.)

thanks,
sven

[Sven S.]

just checked with 1.9.1cvs from 18 Sep 2010 -- no change regarding the case of restricted exogenous regressors (not pushing, just a routine update)

thanks,
sven

[Sven S.]

In April 24's cvs version (2011) this seems to be fixed, so I'm closing this.

(BTW, could be that the added strings in the output are not marked for translation, but that's a different issue.)

thanks,
sven