Avoiding hoizontal lines and crazy shapes when plotting maps in ggplot2

Question

I want a plot of an area, such as Latin America, using the world shape file from IPUMSI...

https://international.ipums.org/international/resources/gis/IPUMSI_world.zip

... I will add some more IPUMS districts later so I really want to use this as my template layer.

I having difficulty with the plot when I add limits through coord_map in ggplot2.

The initial spatial file looks alright

library("ggplot2")
library("raster")

sd0 <- readShapePoly("./IPUMSI_world.shp")
df0 <- fortify(sd0)

ggplot(data = df0, mapping = aes(x = long, y = lat, group = group)) +
  geom_polygon(fill = "black", colour = "black")

When I want to focus on Latin America I get some unwanted horizontal lines:

ggplot(data = df0, mapping = aes(x = long, y = lat, group = group)) +
  geom_polygon(fill = "black", colour = "black") +
  coord_map(xlim = c(-125, -30), ylim = c(-60, 35))

I tried to fix this using the clipPolys function, following the guidance here

library("PBSmapping")
df1 <- df0
names(df1)[c(1,2,6,3)] <- c("X","Y","PID","POS")
df1$PID <- as.numeric(df1$PID)
df2 <- clipPolys(polys = df1, xlim = c(-125, -30), ylim = c(-60, 35), keepExtra = TRUE)
names(df2)[names(df2)=="X"] <- "long" 
names(df2)[names(df2)=="Y"] <- "lat"
names(df2)[names(df2)=="PID"] <- "id"

ggplot(data = df2, mapping = aes(x = long, y = lat, group = group)) +
  geom_polygon(fill = "black", colour = "black")

Not really super happy with this plot either. I thought it was a problem with holes, as in this question, but the suggested solution produces the same plots as above:

gghole <- function(fort){
  poly <- fort[fort$id %in% fort[fort$hole,]$id,]
  hole <- fort[!fort$id %in% fort[fort$hole,]$id,]
  out <- list(poly,hole)
  names(out) <- c('poly','hole')
  return(out)
} 

ggplot(df2, aes(x=long, y=lat, group=group)) +
  geom_polygon(data = gghole(df2)[[1]], fill = "black", colour = "black") +
  geom_polygon(data = gghole(df2)[[2]], fill = "black", colour = "black")

ggplot(df0, aes(x=long, y=lat, group=group)) +
  geom_polygon(data = gghole(df0)[[1]], fill = "black", colour = "black") +
  geom_polygon(data = gghole(df0)[[2]], fill = "black", colour = "black") +
  coord_map(xlim = c(-125, -30), ylim = c(-60, 35))

NJBurgo · Accepted Answer · 2016-08-09 15:22:59Z

8

The other solution is to restrict the view, rather than remove point from the render:

library(ggplot2)
library(maptools)
library(mapproj)

# Maptools dataset
data(wrld_simpl)
world <- fortify(wrld_simpl)

# Same plot, but restrict the view instead of removing points
# allowing the complete render to happen
ggplot(world, mapping = aes(x = long, y = lat, group = group)) +
  geom_polygon(fill = "black", colour = "black") +
  coord_cartesian(xlim = c(-125, -30), ylim = c(-60, 35))

edited Aug 9, 2016 at 15:22

answered Aug 9, 2016 at 15:12

NJBurgo

7793 silver badges8 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

dww Over a year ago

I think this will be slower than setting limits in scale_continuous, because the whole data will be calculated then clipped. scale_continuous will only render the data you want plotted.

NJBurgo Over a year ago

Slower, but you won't get bits of countries missing.

dww Over a year ago

Fair point. So, I guess it depends on the end use which would be preferred. The 'missing' bits of countries are the parts you tell it to omit by setting the limits, which may or may not be what the user wants.

dww · Accepted Answer · 2016-08-09 15:19:28Z

4

You should delete the data outside the plot region. So, use scale_x/y_continuous to set the limits instead of coord_map

ggplot(data = df0, mapping = aes(x = long, y = lat, group = group)) +
  scale_x_continuous(limits = c(-125, -30)) +
  scale_y_continuous(limits = c(-60, 35)) +
  geom_polygon(fill = "black", colour = "black")

edited Aug 9, 2016 at 15:19

answered Aug 9, 2016 at 15:09

dww

31.6k8 gold badges75 silver badges126 bronze badges

Collectives™ on Stack Overflow

Avoiding hoizontal lines and crazy shapes when plotting maps in ggplot2

2 Answers 2

3 Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

3 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related