So after weighting cases, I have been getting this message. To clarify, when on run freq on my weight variable (TRENDWT) there are no missing values and they range btw 3 and 10. Is this stating that the variable it is working with (say for example another variable like GENDER has missing values) is negative/zero/missing? I ask because when running logistic regression (binary), even when I filter out all cases that are negative/sysmis that are being used as variables in the calculation I get this warning. Thank you for any help.