clustering with svy command

[Hannah Lantos]

I'm trying to figure out how to cluster my standard errors while still using the survey command. I'm using an RDS dataset so my weighting (with the svyset command) is based on RDS weights but I'm trying to cluster at a neighborhood level (which had nothing to do with my sampling). Does anyone have any idea how to move forward with this?

It seems like I have 3 options and ideally I'd like to COMBINE them:

1) clustering the standard errors by neighborhood
logistic early_sex_14 xsex ... median_income, robust cluster(ax0NSA)

2) clustering the standard errors with a weight but without taking into account the clustering at the SEED level of the RDS
logistic early_sex_14 xsex ... median_income [pweight=ax0rds_wt], robust cluster(ax0NSA)

3) using the svy command that weights and clusters at the seed level but which doesn't allow me to account for neighborhood clustering
svy: logistic early_sex_14 xsex ... median_income

Any help?

Thanks! Hannah