This repository contains code, plots and data supplementing the article "Modelling admixture across language levels to evaluate deep history claims". The repository is structured as follows:
- the folder code contains the scripts that were run to generate and analyse the results; the files used to run STRUCTURE in the command line are stored in the folder run_structure;
- the folder data contains raw outfiles (outfiles), files with admixture proportions per language (outfiles_cleaned) and files with feature contributions (feature_contributions). Apart from these folders, there is a file with statistics for each run and each K across three language domains (outfile.txt, generated by the script code/collect_outfiles.py), a file with language classifications and coordinates (languages_map.tsv) and a file with feature classifications (feature_set.txt);
- the folder plots contains the plots used in the paper (generated directly by the scripts).