instate estimates state composition associated with Indian surnames using electoral-roll lookups and a calibrated character model. It combines state shares with census mother-tongue shares to derive aggregate language compositions; these are name-pattern estimates, not a person's location or language.
Cite: Gaurav Sood, Rajashekar Chintalapati, and Atul Dhingra. "instate: State and Language Composition from Indian Surnames." https://gsood.com/research/instate-predict-spoken-language-from-last-name/
BibTeX:
@software{soodinstate,
author = {Gaurav Sood and Rajashekar Chintalapati and Atul Dhingra},
title = {{instate: State and Language Composition from Indian Surnames}},
url = {https://github.com/appeler/instate}
}