I have the following data, including mean, std and correlation matrix: data test (type=corr); input _type_$ _name_$ x1 x2 x3 x4 x5 x6 x7 x8; datalines; n . 176 176 176 176 176 176 176 176 mean . 608.65 59.09 0.8075 0.8553 40.75 7.697 94.55 42.56 std . 114.95 38.64 0.3776 0.3554 24.87 2.710 44.48 34.09 corr SP 1.0000 0.8383 0.6149 0.6837 0.7106 0.6891 0.6722 0.7046 corr EFB_POST 0.8383 1.0000 0.5739 0.7181 0.7893 0.7597 0.6851 0.9299 corr SPR 0.6149 0.5739 1.0000 0.6899 0.5478 0.5055 0.4975 0.5639 corr EFR 0.6837 0.7181 0.6899 1.0000 0.7120 0.7109 0.6172 0.6934 corr ECFA 0.7106 0.7893 0.5478 0.7120 1.0000 0.8105 0.7704 0.7546 corr HGC 0.6891 0.7597 0.5055 0.7109 0.8105 1.0000 0.7318 0.7413 corr NLT 0.6722 0.6851 0.4975 0.6172 0.7704 0.7318 1.0000 0.6736 corr EFB_PRE 0.7046 0.9299 0.5639 0.6934 0.7546 0.7413 0.6736 1.0000 ; run; I've tried with the following command to find out the partial correlation, but the result doesn't look right to me. proc proc corr data = test; var x1 x2 partial x3 x4; run; Did I make any mistake there? I am just looking for a better way other than using iml. Thanks in advance!
... View more