msaeDB (version 0.2.1)

datamsaeDBns: Sample Data for Multivariate Small Area Estimation with Difference Benchmarking with clustering

Description

Dataset to simulate difference benchmarking of Multivariate Fay Herriot model for non-sampled area using clustering This data is generated base on multivariate Fay Herriot model by these following steps:

1. Generate explanatory variables X1 and X2. Take \(\mu_{X1}\) = \(\mu_{X1}\) = 10, \(\sigma_{X11}\)=1, \(\sigma_{X2}\)=2, and \(\rho_{x}\)= 1/2. Sampling error e is generated with the following \(\sigma_{e11}\) = 0.15, \(\sigma_{e22}\) = 0.25, \(\sigma_{e33}\) = 0.35, and \(\rho_{e}\) = 1/2. For random effect u, we set \(\sigma_{u11}\)= 0.2, \(\sigma_{u22}\)= 0.6, and \(\sigma_{u33}\)= 1.8. For the weight we generate w1 w2 w3 by set the w1 ~ U(25,30) , w2 ~ U(25,30), w3 ~ U(25,30) Calculate direct estimation Y1 Y2 Y3 where \(Y_{i}\) = \(X * \beta + u_{i} + e_{i}\) cl1 cl2 cl3 were obtained using K-Means clustering from the explanatory variables.

2. Then combine the direct estimations Y1 Y2 Y3, explanatory variables X1 X2, weights w1 w2 w3, and sampling varians covarians v1 v12 v13 v2 v23 v3 in a data frame then named as datamsaeDB

Usage

datamsaeDBns

Format

A data frame with 30 rows and 18 variables:

clY1

cluster information of Y1

clY2

cluster information of Y2

clY3

cluster information of Y3

nonsample

A column with logical values, TRUE if the area is non-sampled

Y1

Direct Estimation of Y1

Y2

Direct Estimation of Y2

Y3

Direct Estimation of Y3

X1

Auxiliary variable of X1

X2

Auxiliary variable of X2

w1

Known proportion of units in small areas of Y1

w2

Known proportion of units in small areas of Y2

w3

Known proportion of units in small areas of Y3

v1

Sampling Variance of Y1

v12

Sampling Covariance of Y1 and Y2

v13

Sampling Covariance of Y1 and Y3

v2

Sampling Variance of Y2

v23

Sampling Covariance of Y2 and Y3

v3

Sampling Variance of Y3