Hello all- So I have the fake sample set below with dates and then values for Field A and B. A is clearly declining over time, B is not. I know how to export out the values such as trendhat so that I can actually produce a graph in ODS later. However, how can I actually export out the p value in this case (for month) to demonstrate that in this case, the "trend" is or is not actually statistically significant for time. I want to take this p value and then translate it into a legend into a graph. Everything else I know how to do. Many Thanks Lawrence ps Sorry for the crude input for time. I really forgot how to manually enter data into a dataset. DATA X; INPUT @2 MONTH_CHAR $ A B; MONTH=input(MONTH_CHAR,mmddyy10.); DATALINES; '01/01/12' 26 26 '02/01/12' 24 12 '03/01/12' 23 26 '04/01/12' 20 17 '05/01/12' 21 25 '06/01/12' 18 19 '07/01/12' 16 28 '08/01/12' 14 20 '09/01/12' 13 18 '10/01/12' 10 20 '11/01/12' 11 14 '12/01/12' 8 8 '01/01/13' 6 20 '02/01/13' 4 22 '03/01/13' 3 6 '04/01/13' 2 25 '05/01/13' 6 28 '06/01/13' 1 22 ; proc autoreg data=X ; model A=month /nlag=2 method=ml; output out=X_A pm=trendhat ;run; proc autoreg data=X ; model B=month /nlag=2 method=ml; output out=X_B pm=trendhat ;run;
... View more
