I want to make some data tables using census 2000 5% sample, that reflect actual ( or very close to actual) number of households in the US. For this I’m not sure how I am to use HHWT. I have one person per household in my list. Weights range from 0 to 2.79 and I’m not sure what these weights are. Are they frequency weights of some sort that I need to manupulate to get the said actual household numbers? or am I supposed to treat them as inverse probabilities? Thank you in advance!