Interpolation at the Beginning and End of a Time Series

Accepted Solution Solved
Reply
New Contributor
Posts: 4
Accepted Solution

Interpolation at the Beginning and End of a Time Series

I am student and a bit of a novice at SAS, so forgive me if this is a simple question.

The EXPAND procedure appears to interpolate missing data IN a time series but ignores missing values at the beginning or end of a series.

How, to I interpolate values at the beginning or end of a series?

Background:

I have over 800 time series that either start later or end sooner than I would like.  While none appear to have missing data in the middle, I am trying to fill in the beginnings and ends to run some analysis.  Thoughts?

Thanks in Advance!


Accepted Solutions
Solution
‎08-29-2013 08:35 PM
Super Contributor
Posts: 307

Re: Interpolation at the Beginning and End of a Time Series

You could use the EXTRAPOLATE option.

View solution in original post


All Replies
Super User
Posts: 10,500

Re: Interpolation at the Beginning and End of a Time Series

I remember doing something this with PROC Arima long ago in a job far far away by asking for predicted values. However depending on the intervals of your time series and number of missing values you'll probably need to look closely to see if results are useable.

Solution
‎08-29-2013 08:35 PM
Super Contributor
Posts: 307

Re: Interpolation at the Beginning and End of a Time Series

You could use the EXTRAPOLATE option.

New Contributor
Posts: 4

Re: Interpolation at the Beginning and End of a Time Series

Any experience on the usability of EXTRAPOLATE?

☑ This topic is SOLVED.

Need further help from the community? Please ask a new question.

Discussion stats
  • 3 replies
  • 200 views
  • 0 likes
  • 3 in conversation