How to generate new variables using -gen- while ignoring missing cases?

[Man Yang]

Hello folks, I need to create a new variable based on the calculation of two existing variables (A and B) but in the meantime, there are missing cases in both A and B. How could I get A/B while letting Stata ignore all the missing cases? I'm aware that -egen- can work with missing option but not -gen-. I don't want to treat missing as 0 as well. Any ideas please? Thanks!

[Joseph Coveney]

I'm not sure that I understand what you're after. generate doesn't treat missing as zero. You're not looking for something like this, are you?

summarize A, meanonly
tempname A
scalar define `A' = r(mean)
summarize B, meaonly
generate double AoverB = `A' / r(mean)

[Rich Goldstein]

Like Joseph, I'm not sure I understand; however,

gen newvar = A/B

will give you the ratio when there are no missing values in an observation and will be missing if either A or B (or both) are missing