In this paper, we propose a Bayesian partition modeling for lifetime data in the presence of a cure fraction by considering a local structure generated by a tessellation which depends on covariates. In this modeling we include information of nominal qualitative variables with more than two...