Hi guys, First of all my apologies if this question is simple, but it was hard to search for similar questions. I have a dataset like this: a b count dif 6 6 6 19 19 13 32 32 13 91 91 59 105 105 14 143 143 38 144 1 172 172 28 199 199 27 272 272 42 326 326 54 356 356 30 Column count equals either a or b, whichever is not missing; while dif is simply the difference between each observation in count and its lag value. My desired output is this: a b count dif 6 6 6 19 19 13 91 91 59 105 105 14 172 172 28 326 326 54 356 356 30 This is to say, I want to keep only observations where a is not missing, while keeping dif value as it was. But once I use where a ne .; the new dif adjusts itself automatically (e.g. row 3 becomes 72 rather than 59). Any suggestions would be greatly appreciated. Cheers Roy ... View more