Readme file

SERIES C  
Applied Statistics

Modelling beyond regression functions: an application of multimodal regression to speed - flow data, by J. Einbeck and G. Tutz
Journal of the Royal Statistical Society, Series C, Applied Statistics, Volume 55 (2006), 461 - 475

The data used in this paper are speed - flow data collected in the 'I-880 FSP Study' (Petty et al., 1996) from a Californian uninterrupted highway ("freeway") having four lanes.

A description of the data is also available at

http://ipa.eecs.berkeley.edu/~pettyk/FSP/Datta/index.html

The data are provided by courtesy of CALIFORNIA PATH, Institute of Transportation Studies, University of California, Berkeley.

The data for lanes 2 and 3 are provided in the files lane2.dat and lane3.dat respectively and can be loaded via

lane2<-read.table("lane2.dat", sep = "\t", header=TRUE)
lane3<-read.table("lane3.dat", sep = "\t", header=TRUE)

Fig. 1 is then obtained by

par(mfrow=c(2,1),cex=0.7)
plot(lane2$flow, lane2$speed)
plot(lane3$flow, lane3$speed)

The modal regression curve in Fig. 3 is obtianed via

modalreg(lane2$flow, lane2$speed, xfix=(1:110)*20, a=100, b=4,
iter= 50, P=2, deg=0, shrink=FALSE)

where xfix is a grid of points where the fitted values are to be calculated, a and b are the bandwidths in x and y- direction respectively, iter is the number of iterations per point and branch, P is the number of starting-points, deg is the degree of the local polynomial and shrink=FALSE means that the regression curves do not get shrunk even for negligible density.

The function modalreg will be included in the R package hdrcde, upcoming version 2.0, or can be obtained from Jochen Einbeck.

References
Petty, K.F., Noeimi, H., Sanwal, K., Rydzewski, D., Skabardonis, A., Varaiya, P., Al-Deek, H. (1996) The freeway service patrol evaluation project: database support programs, and accessibility. Transportation Research Part C: Emerging Technologies, 4, 71-85.

Jochen Einbeck
Department of Mathematics
National University of Ireland
Galway
Republic of Ireland

E-mail: jochen.einbeck@nuigalway.ie

Datasets (einbeck.zip, size - 7KB)

Journals

SERIES A
Statistics in Society

SERIES B
Statistical Methodology

SERIES C
Applied Statistics

SERIES D
The Statistician