A hierarchical model for compositional data analysis

M J Brewer, J A N Filipe, D A Elston, L A Dawson, R W Mayes, C Soulsby, S M Dunn

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

This article introduces a hierarchical model for compositional analysis. Our approach models both source and mixture data simultaneously, and accounts for several different types of variation: these include measurement error on both the mixture and source data; variability in the sample from the source distributions; and variability in the mixing proportions themselves, generally of main interest. The method is an improvement on some existing methods in that estimates of mixing proportions (including their interval estimates) are sure to lie in the range [0, 1]; in addition, it is shown that our model can help in situations where identification of appropriate source data is difficult, especially when we extend our model to include a covariate. We first study the likelihood surface of a base model for a simple example, and then include prior distributions to create a Bayesian model that allows analysis of more complex situations via Markov chain Monte Carlo sampling from the likelihood. Application of the model is illustrated with two examples using real data: one concerning chemical markers in plants, and another on water chemistry.

Original languageEnglish
Pages (from-to)19-34
Number of pages16
JournalJournal of Agricultural, Biological, and Environmental Statistics
Volume10
DOIs
Publication statusPublished - 2005

Keywords

  • Bayesian analysis
  • Markov chain Monte Carlo
  • source apportionment
  • plant alkanes
  • water chemistry
  • N-ALKANES
  • UNCERTAINTY
  • HERBAGE

Cite this

Brewer, M. J., Filipe, J. A. N., Elston, D. A., Dawson, L. A., Mayes, R. W., Soulsby, C., & Dunn, S. M. (2005). A hierarchical model for compositional data analysis. Journal of Agricultural, Biological, and Environmental Statistics, 10, 19-34. https://doi.org/10.1198/108571105X28200

A hierarchical model for compositional data analysis. / Brewer, M J ; Filipe, J A N ; Elston, D A ; Dawson, L A ; Mayes, R W ; Soulsby, C ; Dunn, S M .

In: Journal of Agricultural, Biological, and Environmental Statistics, Vol. 10, 2005, p. 19-34.

Research output: Contribution to journalArticle

Brewer, M J ; Filipe, J A N ; Elston, D A ; Dawson, L A ; Mayes, R W ; Soulsby, C ; Dunn, S M . / A hierarchical model for compositional data analysis. In: Journal of Agricultural, Biological, and Environmental Statistics. 2005 ; Vol. 10. pp. 19-34.
@article{5893eab0e7a14fa19ae661fe1d5b1351,
title = "A hierarchical model for compositional data analysis",
abstract = "This article introduces a hierarchical model for compositional analysis. Our approach models both source and mixture data simultaneously, and accounts for several different types of variation: these include measurement error on both the mixture and source data; variability in the sample from the source distributions; and variability in the mixing proportions themselves, generally of main interest. The method is an improvement on some existing methods in that estimates of mixing proportions (including their interval estimates) are sure to lie in the range [0, 1]; in addition, it is shown that our model can help in situations where identification of appropriate source data is difficult, especially when we extend our model to include a covariate. We first study the likelihood surface of a base model for a simple example, and then include prior distributions to create a Bayesian model that allows analysis of more complex situations via Markov chain Monte Carlo sampling from the likelihood. Application of the model is illustrated with two examples using real data: one concerning chemical markers in plants, and another on water chemistry.",
keywords = "Bayesian analysis, Markov chain Monte Carlo, source apportionment, plant alkanes, water chemistry, N-ALKANES, UNCERTAINTY, HERBAGE",
author = "Brewer, {M J} and Filipe, {J A N} and Elston, {D A} and Dawson, {L A} and Mayes, {R W} and C Soulsby and Dunn, {S M}",
year = "2005",
doi = "10.1198/108571105X28200",
language = "English",
volume = "10",
pages = "19--34",
journal = "Journal of Agricultural, Biological, and Environmental Statistics",
issn = "1085-7117",
publisher = "Springer New York",

}

TY - JOUR

T1 - A hierarchical model for compositional data analysis

AU - Brewer, M J

AU - Filipe, J A N

AU - Elston, D A

AU - Dawson, L A

AU - Mayes, R W

AU - Soulsby, C

AU - Dunn, S M

PY - 2005

Y1 - 2005

N2 - This article introduces a hierarchical model for compositional analysis. Our approach models both source and mixture data simultaneously, and accounts for several different types of variation: these include measurement error on both the mixture and source data; variability in the sample from the source distributions; and variability in the mixing proportions themselves, generally of main interest. The method is an improvement on some existing methods in that estimates of mixing proportions (including their interval estimates) are sure to lie in the range [0, 1]; in addition, it is shown that our model can help in situations where identification of appropriate source data is difficult, especially when we extend our model to include a covariate. We first study the likelihood surface of a base model for a simple example, and then include prior distributions to create a Bayesian model that allows analysis of more complex situations via Markov chain Monte Carlo sampling from the likelihood. Application of the model is illustrated with two examples using real data: one concerning chemical markers in plants, and another on water chemistry.

AB - This article introduces a hierarchical model for compositional analysis. Our approach models both source and mixture data simultaneously, and accounts for several different types of variation: these include measurement error on both the mixture and source data; variability in the sample from the source distributions; and variability in the mixing proportions themselves, generally of main interest. The method is an improvement on some existing methods in that estimates of mixing proportions (including their interval estimates) are sure to lie in the range [0, 1]; in addition, it is shown that our model can help in situations where identification of appropriate source data is difficult, especially when we extend our model to include a covariate. We first study the likelihood surface of a base model for a simple example, and then include prior distributions to create a Bayesian model that allows analysis of more complex situations via Markov chain Monte Carlo sampling from the likelihood. Application of the model is illustrated with two examples using real data: one concerning chemical markers in plants, and another on water chemistry.

KW - Bayesian analysis

KW - Markov chain Monte Carlo

KW - source apportionment

KW - plant alkanes

KW - water chemistry

KW - N-ALKANES

KW - UNCERTAINTY

KW - HERBAGE

U2 - 10.1198/108571105X28200

DO - 10.1198/108571105X28200

M3 - Article

VL - 10

SP - 19

EP - 34

JO - Journal of Agricultural, Biological, and Environmental Statistics

JF - Journal of Agricultural, Biological, and Environmental Statistics

SN - 1085-7117

ER -