WALSpy is a collection of Python functions to explore the World Atlas of Linguistic Structures (WALS). It allows to retrieve information from the database as Pandas objects. It also includes some "experimental" functionality for plotting heatmaps, finding correlations between linguistic features, and predicting unreported properties of a language.
The included raw directory contains the original WALS dataset in csv format. These files (or newer versions of them) are required by WALSpy.
For more information about WALS, visit https://wals.info/. The dataset must be cited as follows.
- Dryer, Matthew S. & Haspelmath, Martin (eds.) 2013. WALS Online (v2020.3) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.7385533
This is a succint review of the functions included in WALSpy. All arguments for these are (i) the id for a language (e.g., 'map' for Mapudungun), (ii) the id for a linguistic feature in WALS (e.g., '33A' for Coding of Nominal Plurality), (iii) the id for a certain feature-value (e.g., '33A-2' for Plural suffix), or (iv) a combination of these.
| Function | Description | Example |
|---|---|---|
| get_features | Retrieves all languages with values for a given linguistic feature | get_feature('81A') |
| get_many_features | Retrieves a dataframe with values for a number n of features | get_many_features(['81A', '82A']) |
| get_totals | Retrieves the number of languages exhibiting values for a given feature | get_totals('81A') |
| features_to_table | Retrieves a table with values for a feature as rows and values for another feature as columns | features_to_table('81A', '84A') |
| search_by_value | Retrieves a set of languages exhibiting a certain value | search_by_value('4A-1') |
| search_many_values | Retrieves a set of languages that coincide in n provided values | search_many_values(['4A-1', '5A-1', '8A-1']) |
| get_genus | Retrieves the genus for a given language | get_genus('map') |
| get_family | Retrieves the family for a given language | get_family('map') |
| get_macroarea | Retrieves the macroarea for a given language | get_macroarea('map') |
| search_features | Given a keyword, the function retrieves a collection of features associated to it. If no keyword is provided, the function retrieves all linguistic features in WALS | search_features('vowel') |
| feature_to_values | Given a feature, the function retrieves a collection of values associated to it. If no feature is provided, the function returns all features with their corresponding values | feature_to_values('17A') |
| list_languages | Retrieves the list of all languages in WALS with their id codes | list_languages() |
| features_to_heatmap | Plots a heatmap showing the relation between two features | features_to_heatmap('83A', '85A') |
Besides these "basic" functions, there are three that should be considered experimental:
| Function | Description | Example |
|---|---|---|
| get_correlation | Checks whether there is a correlation between to two linguistic features; it returns a Cramer's V coefficient and a p-value | get_correlation('83A', '85A') |
| search_correlations | Searches for pairs of features in WALS in linear correlation | search_correlations() |
| predict_value | Predicts the most likely value for a linguistic feature based on a set of values for other features | predict_value('85A', ['81A-3', '86A-2']) |
More details can be found in the docstrings of each function.