Page 1 of 1

Interpolating annual data to quarterly

Posted: Sat Aug 20, 2016 6:23 am
by holyw160
Dear Sir,

I was attempting to derive quarterly interpolated data for housing stock from a time series of annual data. As it is stock data that I am using, I wish to have the final quarter of each year equal to the value from the annual data (end-year) observation, so I presume that under options MAINTAIN=FINAL should be used. The code I have been using is the following:

cal(q) 1970:1
open data "Housing Stock Annual to Qtrly.rat"
data(format=rat) 1970:1 2015:4
set trend = t
source(noecho) disaggregate.src
@disaggregate(factor=4,model=loglin,maintain=final) HSTK / HSTK_q
# trend

However, I get the following error message:

## DLM2. No Observations Produce Valid Output. Check Data and Initial Values The Error Occurred At Location 4054, Line 218 of DISAGGREGATE

It doesn't make sense to use the alternative MAINTAIN=SUM (as it is stock data) so the nearest workable alternative which provides results is MAINTAIN=AVERAGE. This is less than ideal though.

I would appreciate it if you could help identify why I can't get results using MAINTAIN=FINAL.

Thanks in advance.

Re: Interpolating annual data to quarterly

Posted: Sun Aug 21, 2016 2:56 pm
by TomDoan
We fixed the problem with the MAINTAIN=FINAL,MODEL=LOGLIN branch. You can download a newer copy of @DISAGGREGATE.