Library: | metrics |
See also: | adeind wtsder trimper dwade |
Quantlet: | newadeslp | |
Description: | slope estimation of average derivatives |
Usage: | {delta,delta1,dvar} = newadeslp(x,y,h) | |
Input: | ||
x | n x p matrix , the observed explanatory variable | |
y | n x 1 matrix , the observed response variable | |
h | p x 1 vector or scalar , the bandwidth to be used during estimation of the scores | |
Output: | ||
delta | p x 1 vector , the indirect slope estimate | |
delta1 | p x 1 vector , the indirect ADE | |
dvar | p x p matrix , the estimated asymptotic covariance of delta1 |
library("metrics") n = 100 x = normal(n,3) z = 0.2*x[,1] - 0.7*x[,2] + x[,3] eps = normal(n,1) * sqrt(0.5) y = 2 * z^3 + eps h = 5 {delta,delta1,dvar} = newadeslp(x,y,h) delta delta1 dvar
the slope estimator for average derivatives and its asymptotic covariance matrix as described by Stoker in Barnett, Powell, Tauchen, "Nonparametric and Semiparametric Methods in Econometrics and Statistics" (1991) and Turlach, Discussion Paper (1993)