Description
Estimating Peer Effects Using Partial Network Data.
Description
Implements IV-estimator and Bayesian estimator for linear-in-means Spatial Autoregressive (SAR) model (see LeSage, 1997 <doi:10.1177/016001769702000107>; Lee, 2004 <doi:10.1111/j.1468-0262.2004.00558.x>; Bramoullé et al., 2009 <doi:10.1016/j.jeconom.2008.12.021>), while assuming that only a partial information about the network structure is available. Examples are when the adjacency matrix is not fully observed or when only consistent estimation of the network formation model is available (see Boucher and Houndetoungan <https://ahoundetoungan.com/files/Papers/PartialNetwork.pdf>).
README.md
An R package for estimating Peer Effects Using Partial Network Data
The PartialNetwork package includes all functions for the replication of the results in Boucher and Houndetoungan (2023). The exact replication codes are located in the folder test. Below, we also provide detailed examples on how to use the estimators described in the paper.
Installation
CRAN version
PartialNetwork can be directly installed from CRAN.
install.packages("PartialNetwork")
GitHub version
It may be possible that we updated the package without submitting the new version to CRAN. The latest version (but not necessary stable) of PartialNetwork can be installed from this GitHub repos.
remotes::install_github("ahoundetoungan/PartialNetwork", build_vignettes = TRUE)
How to use the PartialNetwork package?
See our vignette in pdf.