| Type: | Package |
| Title: | Bayesian Model to Archaeological Faunal Skeletal Profiles |
| Version: | 1.1.1 |
| Author: | Ana B. Marín-Arroyo [aut], David Ocio [aut], Marco Vidal-Cordasco [cbt], Delphine Vettese [cbt] |
| Maintainer: | Marco Vidal-Cordasco <marcovidalcordasco@gmail.com> |
| Description: | Tool to perform Bayesian inference of carcass processing/transport strategy and bone attrition from archaeofaunal skeletal profiles characterized by percentages of MAU (Minimum Anatomical Units). The approach is based on a generative model for skeletal profiles that replicates the two phases of formation of any faunal assemblage: initial accumulation as a function of human transport strategies and subsequent attrition.Two parameters define this model: 1) the transport preference (alpha), which can take any value between - 1 (mostly axial contribution) and 1 (mostly appendicular contribution) following strategies constructed as a function of butchering efficiency of different anatomical elements and the results of ethnographic studies, and 2) degree of attrition (beta), which can vary between 0 (no attrition) and 10 (maximum attrition) and relates the survivorship of bone elements to their maximum bone density. Starting from uniform prior probability distribution functions of alpha and beta, a Monte Carlo Markov Chain sampling based on a random walk Metropolis-Hasting algorithm is adopted to derive the posterior probability distribution functions, which are then available for interpretation. During this process, the likelihood of obtaining the observed percentages of MAU given a pair of parameter values is estimated by the inverse of the Chi2 statistic, multiplied by the proportion of elements within a 1 percent of the observed value. See Ana B. Marin-Arroyo, David Ocio (2018).<doi:10.1080/08912963.2017.1336620>. |
| License: | GPL-3 |
| Encoding: | UTF-8 |
| Depends: | MASS |
| NeedsCompilation: | no |
| Packaged: | 2024-01-10 10:35:14 UTC; vidalma |
| Repository: | CRAN |
| Date/Publication: | 2024-01-10 10:53:12 UTC |
Bayesian Model to Archaeological Faunal Skeletal Profiles
Description
Tool to perform Bayesian inference of carcass processing/transport strategy and bone attrition from archaeological faunal skeletal profiles characterized by percentages of MAU (Minimum Anatomical Units). The approach is based on a generative model for skeletal profiles that replicates the two phases of formation of any archaeofaunal assemblage: initial accumulation as a function of human transport strategies and subsequent attrition.
Usage
BaSkePro(x)
Arguments
x |
A data frame object with the MAU (Minimum Anatomical Unit) percentage of one species in this specific order: Mandible, Atlas, Axis, Cervical vertebra, Thoracic vertebra, Lumbar vertebra, Rib, Sacrum, Scapula, Humerus, Radius (or radio-ulna), Metacarpal, Pelvis, Femur, Tibia, Metatarsal. |
Details
Two parameters define this model: 1) the transport preference (alpha), which can take any value between - 1 (mostly axial contribution) and 1 (mostly appendicular contribution) following strategies constructed as a function of butchering efficiency of different anatomical elements and the results of ethnographic studies, and 2) degree of attrition (beta), which can vary between 0 (no attrition) and 10 (maximum attrition) and relates the survivorship of bone elements to their maximum bone density. Starting from uniform prior probability distribution functions of alpha and beta, a Monte Carlo Markov Chain sampling based on a random walk Metropolis-Hasting algorithm is adopted to derive the posterior probability distribution functions, which are then available for interpretation. During this process, the likelihood of obtaining the observed percentages of MAU given a pair of parameter values is estimated by the inverse of the Chi2 statistic, multiplied by the proportion of elements within a 1 percent of the observed value.
Value
A table and histograms with posterior probability distribution functions of alpha and beta parameters (the first and fourth quartiles).
Author(s)
Ana B. Marin-Arroyo [aut], David Ocio [aut], Marco Vidal-Cordasco [cbt], Delphine Vettese [cbt]
References
Marín-Arroyo, A. B., & Ocio, D. (2018). Disentangling faunal skeletal profiles. A new probabilistic framework. Historical Biology, 30(6), 720-729. DOI:https://doi.org/10.1080/08912963.2017.1336620
Examples
PMAU <- as.data.frame(c(7,0,0,0,0,0,0,0,0,36,14,0,71,71,57,100))
BaSkePro (PMAU)