Hello, I am working with a larger version of the dataset shown below: yield cellcount htd hya 2.2 260 1 11120101 . 110 1 11120101 5 90 1 11120101 2 400 1 11120101 3 560 1 11120101 4.5 90 2 11120101 2.2 60 2 11120101 3.4 . 2 11120101 1.2 . 2 11120101 1.6 280 2 11120101   I want to set yield and cellcount to missing if the htd or hya have less than 10 records. Edits in htd will affect hya, and the other way around. I use the code below and repeat this a number of times so that I am sure that all htd’s and hya’s have minimum 10 records each in the end. Can you help me with a loop, or something like that, that does this automatically and stops when all htd’s and hya’s have at least 10 records?   Title 'Minimum 10 records per htd - first test'; proc means noprint n data=test maxdec=3; class htd; var yield cellcount; output out=htd_n n=yieldn cellcountn; run; proc sort data=htd_n; by htd; run; proc sort data=test; by htd; run; data test; merge test (in=a) htd_n; by htd; if a=1; if yieldn<10 then yield=.; if cellcountn<10 then cellcount=.; drop _TYPE_ _FREQ_; run; proc means data=test ; var yield cellcount; run; Title 'Minimum 10 records per hya – first test'; proc means noprint n data=test maxdec=3; class hya; var yield cellcount; output out=hya_n n=yieldn cellcountn; run; proc sort data=test; by hya; run; proc sort data=hya_n; by hya; run; data test; merge test (in=a) hya_n; by hya; if a=1; if yieldn<10 then yield=.; if cellcountn<10 then cellcount=.; drop _TYPE_ _FREQ_; run; proc means data=test; var yield cellcount; run;   ... View more