Extensive U.S. data of the deceased and their kin

This data contains metadata for deceased individuals obtained through obituaries in the United States, spanning the period from 2018 to 2022. Additional information is included to facilitate new research and reproduction of prior works. ACSPUMS5Y2022_2025.csv: US Census Bureau estimate for populations by age, marital status, and coresidency. cbsas_kept.csv: the list of CBSAs and a binary value indicated whether they were kept in final analysis (1) or removed (0). The CBSAs are obfuscated. cbsas.csv: original CBSA numbers and names, limited to only the class preserved by the prior file. city_and_kin_excess.csv: specific data necessary to reproduce analysis and plots, provided to preserve anonymity and reduce re-identification risk. decedent_data.csv: extracted information for decedents with corresponding metadata CBSA: Core based statistical area in which the decedent's obituary was printed md5: unique key for the obituary record year: year of death month: month of death age: the decedent's age at death, rounded down to increments of 10 sex: the inferred decedent's gender weight: the weight to reproduce the sampled CBSA mortality of matching demographics and time period decedent_name: a unique index for names provided for the decedent family: a dictionary containing key: name indices utilizing the same indexing as decedent names (decedents can be identified in obituary families using these, and obituaries can be linked by 3rd party names listed in common between them) value: strings indicating the relationship between the decedent and the family member indirect_relationship_types.csv: an encoding matrix given an input of a decedent (row) and a second linked decedent (columns) relationship_lookup.csv: a list of all observed relationship terms observed in obituaries and the corresponding term used in this study.