Dependence in Observational Data

• Individuals are nested in social networks

  • Individual decisions are influenced by their friends.

• Provinces are surrounded by other provinces

  • Provinces mimic one another's policies

• Country-level outcomes are often a result of negotiations with other countries:

  • Economic or environmental policies

Three Mechanisms for Spatial Dependence

• Homophily---similarity in outcomes is endogenous, units are similar because they self-select into the same outcome (e.g., partisan geo-sorting)

• Common exposure---similarity in outcomes is driven by an exogenous factor that affects nearby units (the effect of earthquakes on housing prices)

• Diffusion---nearby units affect each other through learning, imitation, etc (e.g., policy diffusion)

Source: van Weezel S. "On climate and conflict: Precipitation decline and communal conflict in Ethiopia and Kenya." Journal of Peace Research. 2019;56(4):514--528.

Source: Chyzh, Olga V. and R. Urbatsch. 2021. "Bean Counters: The Effect of Soy Tariffs on Change in Republican Vote Share Between the 2016 and 2018 Elections."Journal of Politics 83 (1): 415--419.

What You Need

• Latitude/longitude points for all map boundaries

• Need to know to which boundary/state lat/long points belong

• Need to know the order to connect points within each group

Making Maps

library(tidyverse)
library(mapproj)
library(maps)
library(mapdata)
states <- map_data("state")
head(states)

##        long      lat group order  region subregion
## 1 -87.46201 30.38968     1     1 alabama      <NA>
## 2 -87.48493 30.37249     1     2 alabama      <NA>
## 3 -87.52503 30.37249     1     3 alabama      <NA>
## 4 -87.53076 30.33239     1     4 alabama      <NA>
## 5 -87.57087 30.32665     1     5 alabama      <NA>
## 6 -87.58806 30.32665     1     6 alabama      <NA>

A Basin (Rather Hideous) Map

library(ggplot2)
ggplot() +  geom_path(data=states, aes(x=long, y=lat, group=group),color="black", size=.5)

A Bit Nicer of a Map

#Set theme options:
theme_set(theme_grey() + theme(axis.text=element_blank(),
                               axis.ticks=element_blank(),
                               axis.title.x=element_blank(),
                               axis.title.y=element_blank(),
                               panel.grid.major = element_blank(),
                               panel.grid.minor = element_blank(),
                               panel.border = element_blank(),
                               panel.background = element_blank(),
                               legend.position="none"))
ggplot() +  geom_path(data=states, aes(x=long, y=lat, group=group),color="black", size=.5)+ coord_map()

Polygon instead of Path

ggplot() +  geom_polygon(data=states, aes(x=long, y=lat, group=group),color="black", size=.5)+ coord_map()

Incorporate Information About States

• Add other geographic information (e.g., counties) by adding geometric layers to the plot

• Add non-geographic information by altering the fill color for each state

  • Use geom = "polygon" to treat states as solid shapes to add color

  • Incorporate numeric information using color shade or intensity

  • Incorporate categorical informaion using color hue

Categorical Information Using Hue

If a categorical variable is assigned as the fill color then ggplot will assign different hues for each category.

Let’s load in a state regions dataset:

statereg<- read.csv("./data/statereg.csv")
head(statereg)

##        State StateGroups
## 1 california        West
## 2     nevada        West
## 3     oregon        West
## 4 washington        West
## 5      idaho        West
## 6    montana        West

Join the Data

states.class.map <- left_join(states, statereg, by = c("region" = "State"))
head(states.class.map)

##        long      lat group order  region subregion StateGroups
## 1 -87.46201 30.38968     1     1 alabama      <NA>       South
## 2 -87.48493 30.37249     1     2 alabama      <NA>       South
## 3 -87.52503 30.37249     1     3 alabama      <NA>       South
## 4 -87.53076 30.33239     1     4 alabama      <NA>       South
## 5 -87.57087 30.32665     1     5 alabama      <NA>       South
## 6 -87.58806 30.32665     1     6 alabama      <NA>       South

Plot the Regions

ggplot() +  geom_polygon(data=states.class.map, aes(x=long, y=lat, group=group, fill = StateGroups), colour = I("black"))+ coord_map()+theme(legend.position="bottom")

Map of Canada

library(devtools)
install_github("mccormackandrew/mapcan", build_vignettes = TRUE)
library(mapcan)
canada_map<-mapcan(boundaries = "province", type="standard",territories=TRUE)
head(canada_map)

##      long     lat order  hole piece pr_sgc_code group                pr_english
## 1 2278204 1261783     1 FALSE     1          10  10.1 Newfoundland and Labrador
## 2 2294272 1251745     2 FALSE     1          10  10.1 Newfoundland and Labrador
## 3 2300435 1250534     3 FALSE     1          10  10.1 Newfoundland and Labrador
## 4 2306656 1243299     4 FALSE     1          10  10.1 Newfoundland and Labrador
## 5 2315548 1241630     5 FALSE     1          10  10.1 Newfoundland and Labrador
## 6 2313319 1237013     6 FALSE     1          10  10.1 Newfoundland and Labrador
##                 pr_french pr_alpha
## 1 Terre-Neuve-et-Labrador       NL
## 2 Terre-Neuve-et-Labrador       NL
## 3 Terre-Neuve-et-Labrador       NL
## 4 Terre-Neuve-et-Labrador       NL
## 5 Terre-Neuve-et-Labrador       NL
## 6 Terre-Neuve-et-Labrador       NL

Map of Canada

#Set theme options:
theme_set(theme_grey() + theme(axis.text=element_blank(),
              axis.ticks=element_blank(),
              axis.title.x=element_blank(),
              axis.title.y=element_blank(),
              panel.grid.major = element_blank(),
              panel.grid.minor = element_blank(),
              panel.border = element_blank(),
              panel.background = element_blank(),
          legend.position="none"))
ggplot(canada_map, aes(long, lat, group = group)) +
  geom_polygon(color="black", fill="white")

Map of Canada

Canada Election Results

library(tidyverse)
library(magrittr)
data("federal_election_results") 
federal_election_results %>% as.data.frame() %>% 
  dplyr::filter(election_year=="2015")->electdata
canada_ridings<-mapcan(boundaries = "ridings", type="standard",territories=TRUE)
head(canada_ridings)
canada_ridings %>% left_join(electdata, by="riding_code") %>%
           ggplot(aes(long, lat, group = group, fill=factor(party)))+
  geom_polygon(color="black") +scale_fill_discrete("Party", type="qual") + theme(legend.position="bottom")

Canada Election Results

Your Turn

1. Check out the help file for the map_data function to find out how to make a map of the world.

2. Can you figure out how to remove Antarctica from the map?

3. Use the GDP data you scraped as part of our web-scraping class to shade countries based on their GDP.

Your Turn (Advanced)

1. Read in the animal.csv data:

1. Plot the location of animal sightings on a map of the region
2. On this plot, try to color points by class of animal and/or status of animal
3. Use color to indicate time.