| Title: | Small Area Estimation Non-Parametric Based Nadaraya-Watson Kernel |
|---|---|
| Description: | Propose an area-level, non-parametric regression estimator based on Nadaraya-Watson kernel on small area mean. Adopt a two-stage estimation approach proposed by Prasad and Rao (1990). Mean Squared Error (MSE) estimators are not readily available, so resampling method that called bootstrap is applied. This package are based on the model proposed in Two stage non-parametric approach for small area estimation by Pushpal Mukhopadhyay and Tapabrata Maiti(2004) <http://www.asasrms.org/Proceedings/y2004/files/Jsm2004-000737.pdf>. |
| Authors: | Wicak Surya Hasani[aut, cre], Azka Ubaidillah[aut] |
| Maintainer: | Wicak Surya Hasani <[email protected]> |
| License: | GPL-3 |
| Version: | 0.1.1 |
| Built: | 2025-03-09 05:56:59 UTC |
| Source: | https://github.com/wicaksh/saekernel |
Dataset to Simulate Small Area Estimation Non-Parametric Based Nadaraya-Watson Kernel
This data is generated by these following steps:
Generate explanatory variables Vardir. Vardir ~ abs(N(0, 0.1))
Generate explanatory variables x. x ~ U(min=0, max=1)
Calculate direct estimation y where =
Then combine the direct estimations y, auxiliary variables x, and sampling varians Vardir in a dataframe then named as Data_saekernel
Data_saekernelData_saekernel
A data frame with 100 rows and 3 variables:
Direct Estimation of Y
Auxiliary Variable of X
Sampling Variance of Y
This Function Gives Small Area Estimation Non-Parametric Based Nadaraya-Watson Kernel and Calculates The Bootstrap Mean Squared Error Estimates
mse_saekernel(X, Y, vardir, bandwidth, B = 1000)mse_saekernel(X, Y, vardir, bandwidth, B = 1000)
X |
Auxiliary Variable of X |
Y |
Direct Estimation of Y |
vardir |
Sampling Variances of Direct Estimators |
bandwidth |
The kernel Bandwidth Smoothing Parameter |
B |
Number of Bootstrap. Default is 1000 |
This function returns a list with following objects:
est |
a value of Small Area Estimation Non-Parametric Based Nadaraya-Watson Kernel |
refvar |
Estimated Random Effect Variance |
mse |
Bootstrap Mean Squared Error Estimators of Small Area Estimation Non-Parametric Based Nadaraya-Watson Kernel |
##load dataset data(Data_saekernel) mse_saekernel(X = Data_saekernel$x, Y = Data_saekernel$y, vardir = Data_saekernel$Vardir, bandwidth = 0.04, B = 1000)##load dataset data(Data_saekernel) mse_saekernel(X = Data_saekernel$x, Y = Data_saekernel$y, vardir = Data_saekernel$Vardir, bandwidth = 0.04, B = 1000)
This Function Gives Small Area Estimation Non-Parametric Based Nadaraya-Watson Kernel
saekernel(X, Y, vardir, bandwidth)saekernel(X, Y, vardir, bandwidth)
X |
Auxiliary Variable of X |
Y |
Direct Estimation of Y |
vardir |
Sampling variances of Direct Estimators |
bandwidth |
The kernel Bandwidth Smoothing Parameter |
This function returns a list with following objects:
est |
a value of Small Area Estimation Non-Parametric Based Nadaraya-Watson Kernel |
refvar |
Estimated Random Effect Variance |
##load dataset data(Data_saekernel) saekernel(X = Data_saekernel$x, Y = Data_saekernel$y, vardir = Data_saekernel$Vardir, bandwidth = 0.04)##load dataset data(Data_saekernel) saekernel(X = Data_saekernel$x, Y = Data_saekernel$y, vardir = Data_saekernel$Vardir, bandwidth = 0.04)