Inferences based on regression models for a directional response are usually problematic. This paper presents a Bayesian analysis of a regression model for circular data using the projected normal distribution. Inferences about the model are based on samples from the posterior densities which are obtained using the Gibbs sampler after the introduction of suitable latent variables. The problem of missing data in the response variable is also addressed in this context as is the use of a predictive criterion for model selection. The procedures are illustrated using two simulated datasets a dataset previously analysed in the literature and a real dataset concerning wind directions.
Arnold BC and SenGupta A2006) Recent advances in the analyses of directional data in ecological and environmental sciences. Environmental and Ecological Statistics, 13, 253–56.
2.
Barbieri MM and Berger JO2004) Optimal predictive model selection. The Annals of Statistics, 32, 870–97.
3.
Downs TD and Mardia KV2002) Circular regression. Biometrika, 89, 683–97.
4.
Fisher NI1993Statistical analysis of circular data. Cambridge: Cambridge University Press.
5.
Fisher NI and Lee AJ1992) Regression models for an angular response. Biometrics, 48, 665–77.
6.
Fisher NI, Lewis T and Embleton BJJ1987Statistical analysis of spherical data. Cambridge: University Press.
7.
Geisser S and Eddy WF1979) A predictive approach to model selection. Journal of the American Statistical Association, 74, 153–60.
8.
Gelfand AE and Dey DK1994) Bayesian model choice: asymptotics and exact calculations. Journal of the Royal Statistical Society, Series B, 56, 501–14.
9.
Gelfand AE, Dey DK and Chang H1992Model determination using predictive distributions with implementation via sampling-based methods (with discussion). In Bernardo JM, Berger JO, Dawid AP and Smith AFM (eds.) Bayesian statistics 4. Oxford: Oxford University Press, 147–67.
10.
George BJ and Ghosh K2006) A semiparametric Bayesian model for circular-linear regression. Communications in Statistics (Simulations and Computation), 35, 911–23.
11.
George EI2000) The variable selection problem. Journal of the American Statistical Association, 95, 1304–308.
12.
Gould AL1969) A regression technique for angular variates. Biometrics, 25, 683–700.
13.
Gutiérrez-Peña E1997A Bayesian predictive semiparametric approach to variable selection and model comparison in regression. Bulletin of the International Statistical Institute, Tome LVII. (Proceedings of the 51st Session of the ISI, Invited Papers, Book 1.) Istanbul, Turkey, 17–29.
14.
Jammalamadaka SR and Lund UJ2006) The effect of wind direction on ozone levels: a case study. Environmental and Ecological Statistics, 13, 287–98.
15.
Jammalamadaka SR and SenGupta A2001Topics in circular statistics. Singapore: World Scientific.
16.
Johnson RA and Wehrly TA1978) Some angular-linear distributions and related regression models. Journal of the American Statistical Association, 73, 602–606.
17.
Kato S, Kunio S and Grace SS2008) A circular-circular regression model. Statistica Sinica, 18, 633–45.
18.
Mardia KV and Jupp PE2000Directional statistics. Chichester: Wiley.
19.
Nuñez-Antonio G and Gutiérrez-Peña E2005) A Bayesian analysis of directional data using the projected normal distribution. Journal of Applied Statistics, 32, 995–1001.
20.
Presnell B, Morrisson SP and Littell RC1998) Projected multivariate linear model for directional data. Journal of the American Statistical Association, 93, 1068–77.
21.
R Development Core Team 2009R: A language and environment for statistical computing. Vienna Austria: R Foundation for Statistical Computinghttp://www.R-project.org.
22.
REDMET 2009Atmospheric Monitoring System of Mexico City, http://www.sma.df.gob.mx/sma/ (accessed March 2009).
23.
Rivest LP1997) A decentred predictor for circular-circular regression. Biometrika, 84, 717–26.
24.
Tanner MA and Wong WH1987) The calculation of posterior distributions by data augmentation. Journal of the American Statistical Association, 82, 528–40.
