Windows>> weighted-likelihood>> 返回
项目作者： doug-leasure

项目描述 ：
Proof-of-concept simulations for weighted-likelihood and weighted-precision approaches to deal with weighted survey data for regression analysis.
高级语言： R
项目地址: git://github.com/doug-leasure/weighted-likelihood.git
创建时间： 2019-07-10T16:06:06Z
项目社区：https://github.com/doug-leasure/weighted-likelihood
开源协议：GNU General Public License v3.0
下载


weighted-likelihood

The goal of this repository is to explore options for producing unbiased population estimates using weighted survey data, including a full report describing the methods and findings (Leasure et al 2021). The published report with supplementary data is available from https://doi.org/10.5258/SOTON/WP00706.

This repository includes:

  1. code to simulate population counts in x units
  2. code simulate sampling from the population using weighted or random sampling
  3. code to fit models (stan and jags) to the sample data using weighted-likelihood, weighted-precision, and unweighted approaches
  4. code to assess results with figures and summary statistics
  5. a report to present the findings

Original idea for weighted likelihood approach came from this post by Martyn Plummer on the JAGS forum:
https://sourceforge.net/p/mcmc-jags/discussion/610037/thread/50365933/

Further discussion from Carpenter, Goodrich, and Gelman on the Stan Forum:
https://discourse.mc-stan.org/t/survey-weighted-regression/1654

A relevant paper from Gelman:
https://projecteuclid.org/euclid.ss/1190905511