8

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")

enter image description here

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))

enter image description here

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")

enter image description here

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))
Community
  • 1
  • 1
guyabel
  • 8,014
  • 6
  • 57
  • 86

2 Answers2

8

enter image description hereThe 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))
NJBurgo
  • 779
  • 3
  • 8
  • 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. – dww Aug 09 '16 at 15:17
  • Slower, but you won't get bits of countries missing. – NJBurgo Aug 09 '16 at 15:20
  • 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 Aug 09 '16 at 16:22
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")

enter image description here

dww
  • 30,425
  • 5
  • 68
  • 111