12-12-2011 12:12 PM
I was hoping to get some assistance in understanding a PROC ARIMA error.
I am using a program that was passed along to me that uses PROC ARIMA for seasonal adjustment of data. when I run the program on some sublevels of the dataset being analyzed, I am recieving the following error - "ERROR: The covariance matrix used to compute the Yule-Walker estimates is singular."
First I would like to know exactly what this means - as simply put as possible. And then hopefully get an idea of how to go about fixing it.
In spite of these errors, the program does run fully through and gives me the output I need...but I do worry about having errors in the log and the possibility that the output is not valid.
Thanks in advance for your help.