okay, but I have times series (fund returns) with different start as well as end dates and
therefore missing values might occur at the start at the end or both. if I restrict my
sample to the ones with no missing values, I guess that I will come up with funds that
have a complete history.
kind regards,
Jan
________________________________________
Von: Plus.line MailSystem [cyrus(a)mailer.plusline.de] im Auftrag von Sven Schreiber
[svetosch(a)gmx.net]
Gesendet: Montag, 9. Juli 2012 16:02
An: gretl-users(a)lists.wfu.edu
Betreff: Re: [Gretl-users] Restricting a sample to specific time series
On 07/09/2012 01:44 PM, Jan Tille wrote:
Sven, thank you very much. I will try it. I tried $nobs, but $nobs
always returned 124, which is the length of the total sample period.
No, $nobs should be related to the currently selected sample, and thus
after doing
smpl --no-missing y
it should be different.
-sven
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users