R Package for using Cubic Bezier Spline as a function approximator
especially in modeling latent utility functions.
While Cubic Bezier Splines (CBS) are heavily used in the graphics
software industry, it can also be used as a flexible
function approximation tool given the right constraints. The CBS package
provides a method to calculate the y value
from a x value given an appropriately constrained CBS curve. It then
uses this method to approximate latent utility
functions in intertemporal choice and risky choice data.
CBSr requires rJava package to run
(https://CRAN.R-project.org/package=rJava).
rJava package requires Java development kit (JDK) which needs to be
installed on your computer.
On unix systems, R needs to be reconfigured after JDK installation by
using command ‘R CMD javareconf’.
See rJava webpage under installation section for more details
(http://www.rforge.net/rJava/).
Please use the package on CRAN as it is the latest stable
build.
Package on GITHUB is a development version that is not released yet.
Fun demo video: https://www.youtube.com/watch?v=obR1dpddYow&ab_channel=ArthurSangilLee
Citation: Lee, S., Glaze, C. M., Bradlow, E. T., & Kable, J. W. (2020). Flexible Utility Function Approximation via Cubic Bezier Splines. psychometrika, 85(3), 716-737. https://doi.org/10.1007/s11336-020-09723-4
CRAN: https://CRAN.R-project.org/package=CBSr
GITHUB: https://github.com/sangillee/CBSr