Dietary variability in the Varna Chalcolithic cemeteries – research compendium

This is a compendium of data and R code accompanying our paper at:

Gaydarska, Bisserka, Joe Roe, and Vladimir Slavchev. in press. Dietary variability in the Varna Chalcolithic cemeteries. European Journal of Archaeology.

It describes a cluster analysis of dietary stable isotope measurements from burials at Varna, a Chalcolithic cemetery on the Black Sea coast of modern-day Bulgaria.

We delineated clusters of burials with similar stable isotope ratios using the HDBSCAN algorithm (Campello, Moulavi, and Sander 2013) with $m_{pts}=3$. HDBSCAN (Hierarchical Density-Based Spatial Clustering of Applications with Noise) is an non-parametric clustering and outlier detection algorithm that seeks the ‘most stable’ clusters in a given dataset. In other words, it selects those natural clusters in the data that are least affected by the choice of a particular density or distance threshold. HDBSCAN is well-suited to stable isotope data because it performs well with non-linear clusters, is robust to noise, and doesn’t rely on a pre-specified number of desired clusters (Campello et al. 2015). We applied the modified algorithm suggested by Malzer and Baum (2020), where clustering below a certain threshold distance is ignored. In our case, we selected this threshold to collapse together clusters that were only visible at distances under the maximum measurement error of the isotope ratios (0.03). Clustering was performed with the R package dbscan (Hahsler, Piekenbrock, and Doran 2019); the data and R code to reproduce this analysis is deposited with Zenodo at https://zenodo.org/doi/10.5281/zenodo.11203467 (https://doi.org/10.5281/zenodo.11203468).

Usage

varna_diet_clustering.R contains the R code that performs the cluster analysis. It uses data/varna_human_isotopes.xlsx, which is excerpted from the paper (Table 1). data/varna_human_isotopes.csv contains the extracted and cleaned data used for clustering.

renv.lock records the exact dependency versions used to produce our analysis. You can restore this environment with renv using renv::init().

Citation

Please cite the original paper which this compendium accompanies:

Gaydarska, Bisserka, Joe Roe, and Vladimir Slavchev. in press. Dietary variability in the Varna Chalcolithic cemeteries. European Journal of Archaeology.

License

MIT License

References

Campello, Ricardo J. G. B., Davoud Moulavi, and Joerg Sander. 2013. “Density-Based Clustering Based on Hierarchical Density Estimates.” In Advances in Knowledge Discovery and Data Mining, edited by Jian Pei, Vincent S. Tseng, Longbing Cao, Hiroshi Motoda, and Guandong Xu, 160–72. Lecture Notes in Computer Science. Berlin, Heidelberg: Springer. https://doi.org/10.1007/978-3-642-37456-2_14.

Campello, Ricardo J. G. B., Davoud Moulavi, Arthur Zimek, and Jörg Sander. 2015. “Hierarchical Density Estimates for Data Clustering, Visualization, and Outlier Detection.” ACM Transactions on Knowledge Discovery from Data 10 (1): 5:1–51. https://doi.org/10.1145/2733381.

Hahsler, Michael, Matthew Piekenbrock, and Derek Doran. 2019. “dbscan: Fast Density-Based Clustering with R.” Journal of Statistical Software 91 (1): 1–30. https://doi.org/10.18637/jss.v091.i01.

Malzer, Claudia, and Marcus Baum. 2020. “A Hybrid Approach To Hierarchical Density-based Cluster Selection.” In 2020 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI), 223–28. https://doi.org/10.1109/MFI49285.2020.9235263.