Hello SAS community,   I was reading one of the SAS global proceeding papers related to linear mixed models. That paper suggested to rescaling the data is one of the options to trouble shoot convergence issues in parameter estimation. I'm working on large dataset that consists of multple years and locations, and response variable is yield measured in tons/hectare. Since this dataset requires to estimate a large numbe of parameters, convergence is often a problem to estimate the parameters.  I was thinking to rescale yield values to kgs or some other form.     Any suggestions or comments?   Thank you, RVS ... View more

Hello SAS community, Following is the model that I'm interested to extract blocks of the V matrix using hold option with known or predefined variance parameters. I would like to get V matrix that contains G-side and R-side structure at plot level, i.e.a matrix of 240 x 240 in the following example. I was able to get V matrix for location as subject, but I'm not sure how to change the subject to a plot level to get V with both G and R structures. *Generate numbers; data one;   do har_y = 1 to 2; * two harvest years;    do loc = 1 to 5;  * five locations ;     do blk = 1 to 4; * four blocks;      do cult = 1 to 6;*six cultivars;        yield= 1; * dummy value          output;       end;      end;     end;    end; *Sort proc print; run; proc sort data=one out=two;by har_y loc blk cult; run; proc mixed lognote data=two;   class cult loc har_y blk;   model yield=cult har_y cult*har_y;   random har_y/subject=loc type=csh;   random har_y/subject=loc*cult type=csh;   random har_y/subject=loc*blk type=csh v ;   repeated har_y/subject=loc*blk*cult type=csh r;   parms(6.1387) (3.9508) ( 0.7087) (0.09453) (0.1454) (0.8500) (0.01970) (0.04729) (0) (0.8706) (0.7041) (0.5151)/noiter hold= 1 2 3 4 5 6 7 8 9 10 11 12; *parameters used to get V matrix ods output Tests3 = vcfixed CovParms=vcparms FitStatistics= vcfitstat v=v1 r=r1 ; run; Thank you in advance for your time and consideration. ... View more