CLDF dataset derived from Ugarte et al.'s "NorthPeruLex - A Lexical Dataset of Small Language Families and Isolates from Northern Peru (forthcoming).
If you use these data please cite
- the original source
Ugarte, Carlos and Blum, Frederic and Ingunza, Adriano and Gonzales, Rosa and Peña, Jaime. Forthcoming. NorthPeruLex - A Lexical Dataset of Small Language Families and Isolates from Northern Peru.
- the derived dataset using the DOI of the particular released version you were using
This dataset brings together lexical data from isolates and small language families from northern Peru to investigate their historic relations.
This dataset is licensed under a CC-BY-4.0 license
Conceptlists in Concepticon:
The first step to access all the contents of the dataset is to clone the repository and install all the necessary requirements.
git clone https://github.com/lexibank/northperulex.git
cd northperulex
pip install -e .
This includes all packages used for the conversion to CLDF (Cross-Linguistic Data Formats: https://cldf.clld.org). The NorthPeruLex dataset can also be downloaded directly as a ZIP file directly from this Github repository or from Zenodo (10.5281/zenodo.13269802). If the user wishes to perform the CLDF conversion, they can run the following command:
cldfbench lexibank.makecldf lexibank_northperulex.py --concepticon-version=v3.4.0 --glottolog-version=v5.2.1 --clts-version=v2.3.0
This command uses the cldfbench package (https://pypi.org/project/cldfbench/) with the pylexibank plug-in (https://pypi.org/project/pylexibank/) to automatically convert the data to CLDF using the raw data at the raw folder and the latest version
(at the time of the publication of this dataset) of the references catalogs: Concepticon (https://concepticon.clld.org/), for concept glosses; Glottolog (https://glottolog.org/), for language names; and CLTS (https://clts.clld.org/), for the phonetic transcriptions.
The converted data is located in the cldf folder.
All data in the dataset is stored in tabular (CSV) files. Therefore, it can be read on various platforms and environments and manually inspected.
We provided the user with a analysis\Makefile file that creates a wordlist on a TSV file that can be used to manually inspect the data with the help
of EDICTOR web tool (https://edictor.org/).
To produce the file, please run the following commands:
cd analysis
pip install -r requirements.txt
make wordlist
In addition to yielding the word list file (npl_data.tsv), the Makefile
also runs a script that performs the multiple sequence alignment and an
automatic recognition of sound correspondence patterns. To do so, please
type the following:
make analysisThe result of both processes are stored in the files npl_msaligned
and npl_patterns.tsv.
- Varieties: 35 (linked to 35 different Glottocodes)
- Concepts: 200 (linked to 200 different Concepticon concept sets)
- Lexemes: 4,986
- Sources: 21
- Synonymy: 1.12
- Cognacy: 4,986 cognates in 3,660 cognate sets (2,905 singletons)
- Cognate Diversity: 0.72
- Invalid lexemes: 0
- Tokens: 29,488
- Segments: 185 (0 BIPA errors, 0 CLTS sound class errors, 184 CLTS modified)
- Inventory size (avg): 30.00
| Name | GitHub user | Description | Role |
|---|---|---|---|
| Carlos Ugarte | @CMUgarte | Data collector, CLDF conversion and annotation | Author, Editor |
| Frederic Blum | @FredericBlum | CLDF conversion and annotation | Author, Editor |
| Adriano Ingunza | @BadBatched | Data collector and annotation | Author |
| Rosa Gonzales | @rosalgm | Data collector and annotation | Author |
| Jaime Peña | @JaimePenat | Data collector and annotation | Author |
The following CLDF datasets are available in cldf:
- CLDF Wordlist at cldf/cldf-metadata.json