provenance bundles a number of established statistical methods to facilitate the visual interpretation of large datasets in sedimentary geology. Includes functionality for adaptive kernel density estimation, principal component analysis, correspondence analysis, multidimensional scaling, generalised procrustes analysis and individual differences scaling using a variety of dissimilarity measures. Univariate provenance proxies, such as single-grain ages or (isotopic) compositions are compared with the Kolmogorov-Smirnov, Kuiper or Sircombe-Hazelton L2 distances. Categorical provenance proxies such as chemical compositions are compared with the Aitchison and Bray-Curtis distances, and point-counting data with the chi-square distance. Also included are tools to plot compositional and point-counting data on ternary diagrams and point-counting data on radial plots, to calculate the sample size required for specified levels of statistical precision, and to assess the effects of hydraulic sorting on detrital compositions. Includes an intuitive query-based user interface for users who are not proficient in R..


You must have R installed on your system (see Additionally, to install provenance from Github, you also need the devtools package. This can be installed by typing the following code at the R command line prompt:



The most recent stable version of provenance is available from CRAN at and can be installed on your system as follows:


Alternatively, to install the current development version of provenance from Github, type:


Further information


Vermeesch, P., Resentini, A. and Garzanti, E., 2016, An R package for statistical provenance analysis, Sedimentary Geology, 336, 14-25

Vermeesch, P., 2018, Statistical models for point-counting data. Earth and Planetary Science Letters 501, 1-7


Pieter Vermeesch


This project is licensed under the GPL-3 License