On Mon, 15 Mar 2021, Sven Schreiber wrote:
when trying out the relatively new temporal disaggregation with the
quarterly AWM.gdt dataset in the GUI, I "always" get the error message
that the series does not appear to be suitable for that. For example
series CAN or GCN.
Is it because it is already quarterly and conversion to monthly isn't
supported? I confess I haven't read the entire chapter 9 of the guide on
the topic yet...
With quarterly AWM loaded, you can't create a monthly series. But
you can get CAN (e.g.) as a monthly column vector:
<hansl>
open AWM.gdt
smpl ok(CAN) --restrict
matrix CAN12 = tdisagg(CAN, ,3, _(verbose=1))
</hansl>
The "does not seem suitable" message needs revision, I guess. A
series that's "suitable" for in-place disaggregation would be one
that is of lower frequency than the current dataset, this being
indicated by one or other of these patterns:
* Repetition of low-frequency value per sub-period (e.g. an annual
value appearing in each of the 4 quarters).
* Value given for only the first sub-period (e.g. annual value
appears only in Q1, with NAs otherwise).
Allin
_______________________________________________
Gretl-users mailing list -- gretl-users@gretlml.univpm.it
To unsubscribe send an email to gretl-users-le...@gretlml.univpm.it
Website:
https://gretlml.univpm.it/postorius/lists/gretl-users.gretlml.univpm.it/
