I have a question for complex survey analysis. I want to get racial breakdown by occupation using IPUMS ACS data. The problem is that I recreate my own race variable using both