Using imputation before bartcause

[njawadekar]

To handle missing data in covariates that will be ultimately used in BART, is there a particular imputation technique (if any) or a standard protocol that you recommend we use before running causal inference with bartcause?