I do not understand your question. The numbers 9e100 and 9d100 are the same number. They both mean 9 times 10 raised to the 100th power.  In the old days of Fortran they used the D instead of the E to mean to use double precision (8 byte) floating point instead of single precision (4 bytes).  But in SAS all numbers are stored in 8 byte floating point.   If the field is an ACCOUNT number then you should probably be treating it as a character string in SAS. You will never need to take the mean of an ACCOUNT number.  Also there is a limit to the number of digits that an 8 byte floating point number can represent exactly.   1 data _null_; 2 x=constant('exactint'); 3 put x comma24. ; 4 run; 9,007,199,254,740,992   ... View more