Subscribe to DSC Newsletter

Determining Seasonality using PROC X11 in SAS

Determining seasonality, deseasonlizing seasonal data, getting trend indices, getting seasonality indices have always been a boring nightmare for Analysts when it comes to analysing percentage increase in some parameter like sales, visits, frauds etc., negating the effects due to seasonality. All the above mentioned so called "NIGHTMARES"can be easily handled with the powerful but less used SAS procedure called PROC X11.

PROC X11 works as below,
It splits the actual data into three parts
1. The Seasonality Component (Seasonal peaks, dips etc.,)
2. The Trend Component (Positive/Negative slope in the data plot)
3. Irregular/Random component (which cannot be explained by trend and seasonlity)

The Proc X11 output contains the following series ,
b1 represents the original data
d10 represents the seasonality component in the data
d11 represents the deseasonlised data, i.e., the data made free of seasonal peaks/dips
d12 represents the trend component
and lastly d13 represents the random/irregular component in the data

1. Prevents the analysts from spending lots of time by writing long formulae in Excel/SAS to get rid of seasonality
2. Also gives a neat continuous variable expalining seasonality, trend etc.,
3. Very powerful trend/seasonlity analysis technique
4. Very useful for predicting trends/seasonlity for better accuracy of forecasts

1. Can be used only if there is minimum three years of data (either quarterly/monthly)

SAS Code to run PROC X11:

proc x11 data=dataset;
monthly date=date;/*needs to be in date9. format*/
tables d10 d11 d12 d13;
output out=output_dataset b1=originald10=seasonald11=adjustedd12=trendd13=irreg;
var variable;/*variable is the variable to be analysed*/

Views: 5735

Tags: d10, d11, d12, d13, irregularity, proc, procx11, sas, seasonality, trend, More…x11


You need to be a member of AnalyticBridge to add comments!

Join AnalyticBridge

On Data Science Central

© 2020   TechTarget, Inc.   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service