SAS的varmax语句里提供该检验，SAS/IML也可以做

proc Varmax data=work.shenzheng;
model uplinecfone downline /p=2;
causal group1=(uplinecfone) group2=(downline);/*H0:X2不是X1的Granger原因*/
run;
p=2或者p=1什么意思？

为什么结果没有因果检验
data shen;
set  work.shen;
run;
/*ARMA估计
proc arima；
identify var=x nlag=12  ;run;
proc arima; identify var=x(1,12) nlag=12;run;
proc arima;
identify var=x(1,12) nlag=12 minic p=(0,5) q=(0,5);BIC值
estimate p=4 q=2  method=cls;*/


proc Varmax data=work.shen;                     /*格兰杰因果检验*/
model  dupline ddownline   /p=2;
causal    group1=(dupline)  group2=(ddownline);
causal  group1=(ddownline) group2=(pp);run;   /*H0:X2不是X1的Granger原因*/
proc Varmax data=work.shen;
model  dupline pp       /p=2;
causal    group1=(dupline)  group2=(pp);
causal  group1=(pp)  group2=(dupline);
run;
proc Varmax data=work.shen;
model  ddownline  pp     /p=2;
causal   group1=(pp)  group2=(ddownline);
causal  group1=(ddownline)    group2=(pp);
run;
proc Varmax data=work.shen;
model  dupline  dbody    /p=2;
causal  group1=(dupline)group2=(dbody);
causal group1=(dbody) group2=(dupline);
run;
proc Varmax data=work.shen;
model  ddownline  dbody   /p=2;
causal  group1=(dbody) group2=(ddownline);
causal group1=(ddownline) group2=(pp);
run;
proc Varmax data=work.shen;
model  pp  dbody    /p=2;
causal  group1=(dbody) group2=(pp);
causal group1=(pp)group2=(dbody);
run;

shen数据集里包含了所有变量