> I am a bit puzzled by the grelt syntax. Why can I not use (s is a
> series)
>
> print s[1]
>
> but I can use
>
> scalar z = s[1]
>
> print z
Although it may seem so, this doesn't really have to do with accessing series
elements. The 'print' command is limited to work on one pre-specified
object.
The alternative for you is to use 'eval' to evaluate a general hansl
expression:
eval s[1]
> ??? It cost me some time to get it that actually accessing single
> elements works as it is written in hansl, simply because I did not
> understand the error msg from using the print command
I can confirm that the error message is strange: I get "invalid observation
index 1", which is certainly misleading.
> or leading ? to check my understanding of syntax...). Also, in hansl it
> is said that one may use
>
> lagged_s =s[-1]
>
> to produce a lagged version of the series. But to no avail (at least
> not, if there are some NA included). Wheras the (undocumented?)
>
> lagged_s = s(-1) works
Where is it said that you should use square brackets instead of round ones?
(honest question) In my memory it has always been s(-1).
thanks,
sven
You are right, the -1 +1 is explained to be needed in round squares ALWAYS (Hansl Manual
p. 45 was where I got confused, I checked back). But the [] not working with print/?
really got me struck...