On Mon, 11 Jul 2011, Mariusz Doszyń wrote: > Is there a command in gretl that makes possible to omit observations with > missing values while using estimation procedures? I know that it can be also > done by using such commands as misszero, if, zeromiss and so on but maybe > there is a simpler way?
Missing observations are automatically skipped when that is feasible -- see section 4.6 of the Gretl User's Guide. If you want to handle cases where gretl cannot work around missing values you can use the --no-missing form of the "smpl" command to create an OK subset of your data; see the command reference for "smpl". Allin Cottrell