In arnoblalam/risk_sharing: Risk Sharing Networks in Tanzania

knitr::opts_chunk$set(echo = TRUE)

Purpose

The purpose of this vignette is to check the Nyakatoke village data provided in (source) and compare descriptive statistics with what is stated in published works.

Data loading/saving

This part of the code is simply loads the raw data in the extdata folder and saves it as an R data set

baseLoc <- system.file(package="riskSharing") 
extPath <- file.path(baseLoc, "extdata")

nyakatoke <- read.csv(file.path(extPath, file.path("tanzania_data.csv")))
save(nyakatoke, file = "../data/nyakatoke.RData", compress = "xz")

Ananlysis

We load the igraph package and reload the data

library(riskSharing)
data("nyakatoke")

Recreate DeWeerdt (2006)

Let's create an undirected graph from the data. In this case, we say two households share a link if either household reported the other.

library(igraph)
g <- graph_from_data_frame(nyakatoke[nyakatoke$willingness_link1 == 1 |
                                nyakatoke$willingness_link2,], directed = FALSE)
g <- simplify(g)

library(xtable)
g.order <- length(igraph::V(g))
g.size <- length(igraph::E(g))
g.mean.degree <- mean(degree(g))
g.min.degree <- min(degree(g))
g.max.degree <- max(degree(g))
g.med.degree <- median(degree(g))
g.distances <- distance_table(g)$res
names(g.distances) <- 1:length(g.distances)
summary.stats <- data.frame(g.mean.degree, g.med.degree, g.min.degree, 
                            g.max.degree)
xt <- xtable(summary.stats, auto = TRUE, digits = 2)
names(xt) <- c("Mean", "Median", "Min", "Max")
print(xt, type = "html")