This package provides functions for the age-at-death estimation of nonadult human skeletal remains based on dental wear. The method was developed using a Dutch post-Medieval reference population; however, the package also allows the inclusion of other (more appropriate) reference samples. The main function allows the age-at-death estimation of nonadults based on a single wear score calculated from all available teeth, and additional functions allow the user to upload and evaluate their own reference sample, from which to base the age-at-death estimations. Users who are not familiar with the R-environment can also take advantage of a user-friendly graphical user interface (GUI).
Installation
- To install BAMSAUR, R (>= 3.6.0) must first be installed (https://cran.r-project.org/).
#2. Open R, and install the package "devtools" by entering the following into the R-console:
install.packages("devtools")
#3. The BAMSAUR package can then be downloaded and installed:
devtools::install_github("bbartholdy/BAMSAUR")
library(BAMSAUR)Usage
The BAMSAUR package uses the following R-packages:
c("AICcmodavg", "earth", "ggfortify", "ggplot2", "shiny")The method was developed on Middenbeemster, a 19th century Dutch skeletal collection containing individuals with documented age-at-death. The method uses a 9-stage dental wear scoring system, applicable to the entire deciduous and permanent (excluding third molars) dentitions. The average wear scores were used to develop a quadratic regression model from which to base age-at-death predictions. Age ranges are calculated using 68% prediction intervals (PIs). Users can also elect to use linear and cubic regression, or multivariate adaptive regression splines (MARS) for age-at-death predictions. Additional functions can be used to evaluate potential reference samples by developing linear, quadratic, and cubic regression models, as well as a MARS model (using the ‘earth’ package), and provides information on the developed models such as r-squared values, accuracies from leave-one-out cross validations (LOOCV), predictive residual error sum of squares (PRESS), Akaike and Bayesian information criterion (AIC and BIC), and precision (based on the average size of predicted age ranges). It also returns plots (using ‘ggplot2’) for each of the regression models.
Functions
BAMSAUR age-at-death estimation
BAMSAUR(wear, data = NULL, pop = "MB11", rank = 2, interval = "prediction", level = 0.68, mars.int = TRUE, ...)
The main function which can be used to predict age-at-death based on dental wear by entering an obtained wear score, following the dental wear method outlined in the main article (Bartholdy et al., under review). The Middenbeemster data is built-in, so if pop = “MB11” is chosen, the data input is not required. If pop = “other”, a data frame with a column for age and a second column for wear scores can be inserted. The function returns the input wear score, the age prediction, age range as +- years, and lower and upper intervals. The default method is a quadratic (rank = 2) model with 68%PIs, but “confidence” intervals and any level between 0 and 1 (0 - 100%), not inclusive, can be chosen. The input ‘mars.int’ allows the user to apply MARS-sized intervals to the estimates from simple regression models. The recommended setting for estimations based on the MB11 reference sample is TRUE, as this reference sample contains some individuals without documented age, and the MARS-sized age intervals can better account for this error. If ‘mars.int = FALSE’, regular confidence/prediction intervals are provided.
The BAMSAUR.mars function serves the same purpose as the BAMSAUR function, but uses a MARS model for the age-at-death estimation. Only “prediction” intervals are supported for MARS models. The default number of cross validations (ncross) is 3 to reduce computation time.
Examples:
#The following example calculates the ages-at-death with MARS 68%PIs of 10 random wear scores
wear <- runif(10,0,16)
BAMSAUR(wear)
#Ages-at-death using a linear model and 95%CIs
BAMSAUR(wear, rank = 1, interval = "confidence", level = 0.95, mars.int = F)
#Age-at-death calculation using MARS
wear <- runif(10,0,16)
BAMSAUR.mars(wear)
#Age-at-death calculation using 90%PIs and a custom dataset
data <- MBsimple[1:25]
BAMSAUR.mars(wear, data = data, pop = "other", level = 0.9)BAMSAUR best-fit function
“It’s not the sample size that matters, it’s the goodness-of-fit!”
This function can be used to evaluate a dataset to determine its feasibility as a reference sample for estimating age-at-death in the main BAMSAUR function. The data input must contain one column with known ages-at-death (whether osteologically estimated or documented age) and a second column with the associated wear scores. The function returns a summary of the developed linear, quadratic, and cubic regression models, as well as a MARS model. It also returns values to indicate goodness-of-fit of the various models, including r-squared values, accuracies from the LOOCV, PRESS, AIC (corrected for small sample sizes), BIC, and average age range (to indicate precision), as well as diagnostic plots.
Examples:
#the following expression evaluates the sample with 68% prediction intervals as age ranges
MBex <- BAMSAUR.bff(MBsimple)
#the following code accesses the quadratic plot
MBex$quad.plotBAMSAUR shiny app
This launches the BAMSAUR shiny app, which has similar functionality as the individual functions.
Example:
##Not run: runBAM()BAMSAUR data
The BAMSAUR data contains two datasets: MBsimple and MBdata. MBsimple contains a column with known ages-at-death and associated wear scores for the Middenbeemster sample, which is built-in to the BAMSAUR and BAMSAUR.mars functions. MBdata contains all the wear scores for the individual teeth of each specimen that was used for the analysis, and is included for sharing and reproducibility purposes.
References
Bartholdy, B.P., Hoogland, M.L.P, and Waters-Rist, A., under review. How Old Are You Now? A new ageing method for nonadults based on dental wear. International Journal of Osteoarchaeology. doi: 10.1002/oa.2758.
Hadley Wickham, Jim Hester and Winston Chang (2018). devtools: Tools to Make Developing R Packages Easier. R package version 1.13.6. https://CRAN.R-project.org/package=devtools
Hadley Wickham, Peter Danenberg and Manuel Eugster (2018). roxygen2: In-Line Documentation for R. R package version 6.1.0. https://CRAN.R-project.org/package=roxygen2
JJ Allaire, Jeffrey Horner, Vicent Marti and Natacha Porte (2017). markdown: ‘Markdown’ Rendering for R. R package version 0.8. https://CRAN.R-project.org/package=markdown
Marc J. Mazerolle (2017) AICcmodavg: Model selection and multimodel inference based on (Q)AIC(c). R package version 2.1-1. https://cran.r-project.org/package=AICcmodavg.
Milborrow, S. Derived from mda:mars by Trevor Hastie and Rob Tibshirani. Uses Alan Miller’s Fortran utilities with Thomas Lumley’s leaps wrapper. (2019). earth: Multivariate adaptive regression splines. R package version 4.7.0. https://CRAN.R-project.org/package=earth publication
R Core Team (2018). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.
Wickham, H. (2015). R packages: organize, test, document, and share your code. O’Reilly Media, Inc.
Wickham, H. (2016). ggplot2: Elegant graphics for data analysis. New York: Springer-Verlag.
Winston Chang, Joe Cheng, JJ Allaire, Yihui Xie and Jonathan McPherson (2018). shiny: Web Application Framework for R. R package version 1.2.0. https://CRAN.R-project.org/package=shiny
Yuan Tang, Masaaki Horikoshi, and Wenxuan Li. “ggfortify: Unified Interface to Visualize Statistical Result of Popular R Packages.” The R Journal 8.2 (2016): 478-489.
