* Description of scripts and datasets

** Datasets:

*** df-final.csv: Dataset with the columns
- Language_ID : id as found in WALS
- Glottocode : corresponding glottocode
- name_language : name of the langauge
- latitude : latitude information of the langauge
- longitude : longitude information of the langauge
- name_micro_family : the node directly above the language in the glottolog tree
- affixation : Affix position. Values are: Strongly suffixing, Weakly suffixing, Equal prefixing and suffixing, weakly prefixing, Strongly suffixing
- family : family information about the language (for languages without family information we take the language itself as a family)
- macroarea : macroarea information about the language
- vo_order : Verb-Object order. Values are: OV, No dominant order, VO.
- lat : Latitude centered by macroarea
- long : Longitude centered by macroarea

*** df-phylo.rds : Phylogenetic tree with the languages in df-final.csv

** Scripts

*** model-main.R: script to build the models with the original dataset. The models included are:

- model_main : main proposed model.
- model_main_prior : for checking the priors (not described in the paper but used for evaluating our prior choice).
- model_no_areal : model without any areal effects but with phylogenetic effects.
- model_no_areal_2 : model without any areal effects but with family effects.
- model_novo : model with areal and phylogenetic effects but without verb-object order as covariate.
- model_no_phylo : model with areal effect and family effect but no phylogenetic effect.
- model_no_fam : model with areal effect.
- model_trad : hierarchical model with family effect and macroarea effect.
- model_nc : model without controls.

*** model-oversample.R : script to build the models with oversampling the original dataset. The models included are:

Oversampling of Indo-European languages:

- model_over_ie_all : main model with areal and phylogenetic controls.
- model_over_ie_no_controls : model with no controls.
- model_over_ie_traditional : hierarchical model with family effects and macroarea effects.

Oversampling of South American languages:

- model_over_sa_all : main model with areal and phylogenetic controls.
- model_over_sa_no_controls : model with no controls.
- model_over_sa_traditional : hierarchical model with family effects and macroarea effects.

*** analyze-aff.R : script to create the plots and tables in the paper.

*** generate-phylonegy.R : script to build the phylogenetic tree.