Kinkreg: Kink regression estimation

View source: R/kink.R

KinkregR Documentation

Kink regression estimation

Description

The least sqaure estimation of kink regression model

Usage

Kinkreg(y,x,z,Call="optim")

Arguments

y

the vector of dependent variable

x

the matrix of regime-dependent independent variable

z

the matrix of exogenous variable

Call

if Call="optim" == optimization if Call="residuals" == retrun the residul series

Details

The kink regression model is the nonlinear model in which the continous function is change at the kink point.

Value

Result

the parameter estimates, standard error, t-stat and p-value

RSS

Residual sum Sqaures

Author(s)

Woraphon Yamaka

References

Hansen, B. E. (2017). Regression kink with an unknown threshold. Journal of Business & Economic Statistics, 35(2), 228-240.

Sriboochitta, S., Yamaka, W., Maneejuk, P., & Pastpipatkul, P. (2017). A generalized information theoretical approach to non-linear time series model. In Robustness in Econometrics (pp. 333-348). Springer, Cham.

Maneejuk, P., Yamaka, W., & Sriboonchitta, S. (2019). Does the Kuznets curve exist in Thailand? A two decades’ perspective (1993–2015). Annals of Operations Research, 1-32.

Maneejuk, P., & Yamaka, W. (2020). An analysis of the impacts of telecommunications technology and innovation on economic growth. Telecommunications Policy, 44(10), 102038.

Maneejuk, P., & Yamaka, W. (2021). The Impact of Higher Education on Economic Growth in ASEAN-5 Countries. Sustainability, 13(2), 520.

Examples

#### Example Simulation data

set.seed(111)                 # Set seed for reproducibility
k = 1                         #number of dependent variable
n = 500                       #number of observation
r1=1.5                        # Kink parameter
x = matrix(rnorm(n*k,r1,2),ncol=k)
e = rnorm(n,0,sd=1)
x1 = cbind(neg.part(x-r1),pos.part(x-r1))
z = cbind(rnorm(n,0,sd=1))
y=0.5+(0.5*x1[,1])-(1*x1[,2])+0.5*z+e
plot(x,y, col="red", lwd=2)

Kinkreg(y,x,z,Call="optim")

woraphonyamaka/Kinkreg documentation built on Aug. 15, 2022, 4:17 a.m.