I have a dataset from March 2006 to January 2018.
Using proc timeseries, I am trying to identify the trend from a time series dataset.
ods graphics on;
proc timeseries data = all_posters_ext plots = (series corr decomp) outdecomp = all_posters_ext_trend;
id date interval = month accumulate = avg;
var nbr_poster;
decomp ORIG TCC / lambda = 10000;
run;
However, the TCC variable given by that code give empty observations from March 2006 to August 2006 and from August 2017 to January 2018.
Is it possible to expand the TCC to cover the whole time period analyzed (March 2006 to January 2018)?
Thank you in advance
The moving averages are centered around the observations. So there will be some missing values at the beginning and end. You can find the formulas here (look at the end of the page)
However, the window of data that is missing the TCC component at the end is too wide for your data. That's why I was suggesting to check if there are missing values.
If you want estimates for the TC component for all the range of the data, and into the future, I suggest you look at models to forecast it. For example PROC ESM and PROC UCM, or PROC X12 and X13.
The TCC(trend-cycle) component is a moving average of the data. How do your data look in those periods? Any missing values?
There are no missing values.
I understand that the beginning of the period might cause problems as moving average need historical data. However, no results are also given for the end of the timeline.
The moving averages are centered around the observations. So there will be some missing values at the beginning and end. You can find the formulas here (look at the end of the page)
However, the window of data that is missing the TCC component at the end is too wide for your data. That's why I was suggesting to check if there are missing values.
If you want estimates for the TC component for all the range of the data, and into the future, I suggest you look at models to forecast it. For example PROC ESM and PROC UCM, or PROC X12 and X13.
Registration is open! SAS is returning to Vegas for an AI and analytics experience like no other! Whether you're an executive, manager, end user or SAS partner, SAS Innovate is designed for everyone on your team. Register for just $495 by 12/31/2023.
If you are interested in speaking, there is still time to submit a session idea. More details are posted on the website.
Learn how to run multiple linear regression models with and without interactions, presented by SAS user Alex Chaplin.
Find more tutorials on the SAS Users YouTube channel.