Re: [Gretl-users] problem with system instrumental variables ==> runs OK

Hello gretl listers ! Greetings from Paris ! First, thank you Allin for the quick fix, you save my day (and maybe my week !) Just installed winsnapshot and tried the 500 observations sample script with the tsls syntax command. All runs smoothly and perfectly ! Once again thank you for the good job. I will try later my that day the estimation on the whole sample . I observed that the TSLS type syntax is available for the following sytem estimators ('method=xxx' switch) : TSLS ; 3SLS but not for SUR and OLS (I did not try the others). That's perfect as I can do my special Heiein-Wessels 2 step estimation. By the way, i am currently working intensively with system estimation on microdata, and this is the reason of my modest contribution. I am also fond of GRETL and try to do most of my work the jewel, encountering these little problems. Again, thanx, and continue the good job ! sincerely Franck