In UNFAOstatistics/gisfao: UNFAO GIS R tools

library(knitr)
opts_chunk$set(list(echo=TRUE,eval=TRUE,cache=FALSE,warning=FALSE,message=FALSE,fig.height=6,fig.width=8,fig.path = "../figure/"))

Intro

In this experiment we will download .csv data from FAOSTAT and create a simple choropleth maps using ggplot2.

Downloading data from FAOSTAT

In this example we are not using the FAOSTAT API for downloading the data, but we download the Live Animals data through bulk download link as below.

download.file("http://faostat3.fao.org/faostat-bulkdownloads/Production_Livestock_E_All_Data_(Norm).zip",destfile = "bulk.zip")
unzip("bulk.zip")
dat <- read.csv("Production_Livestock_E_All_Data_(Norm).csv", stringsAsFactors = FALSE)

Once downloaded we use the dplyr-package to subset the number of cattle for years 2000, 2005 and 2010

library(dplyr)
d <- dat %>% 
  filter(Item == "Cattle", Unit == "Head", Year %in% c(2000,2005,2010)) %>% # filtering the rows
  select(Country.Code,Year,Value) %>% # selecting the columns
  rename(FAOST_CODE = Country.Code) # renaming the var
knitr::kable(head(d)) # showing the first 6 rows of the data

Then we classify the data manually also using dplyr-package

library(dplyr)
d <- d %>% mutate(cat_man = cut(Value, breaks=c(0, 97040, 1452000, 9149000, 1453000000), include.lowest=TRUE), # cut manually
                  cat_int = ggplot2::cut_interval(Value,n = 5), # cut into 5 equally interval
                  cat_num = ggplot2::cut_number(Value, n = 5)) # cut into 
knitr::kable(head(d)) # showing the first 6 rows of the data

Manipulating the spatial data

For simplicity we will first fortify the SpatialPolygonDataFrame into data.frame and merge it with the attribute data from FAOSTAT only thereafter. However, I strongly advice to use approach in the gisfao_vignette.Rmd that uses the proper spatial data types and methods. It is good to remember tha you can't go back into spatial data classes from normal data.frame meaning that the following operations are irreversible!

library(gisfao)
library(rgdal)
library(ggplot2)
shape <- spTransform(fao_world, CRS("+proj=robin"))
map.df <- fortify(shape)

Merge the map.df with the FAOSTAT data

map.df <- merge(map.df,d,by.x="id", by.y="FAOST_CODE",all=TRUE)
map.df <- arrange(map.df, order)

Create maps

Using each of three category variables and year 2010

Using cut

plot_data <- map.df %>% filter(Year == 2010)

p <- ggplot()
p <- p + geom_polygon(data=map.df,aes(long,lat,group=group,fill=cat_man), 
                      color="white", size=.5)
p <- p + theme_minimal()
p

Using cut_interval

plot_data <- map.df %>% filter(Year == 2010)

p <- ggplot()
p <- p + geom_polygon(data=map.df,aes(long,lat,group=group,fill=cat_int), 
                      color="white", size=.5)
p <- p + theme_minimal()
p

Using cut_number

plot_data <- map.df %>% filter(Year == 2010)

p <- ggplot()
p <- p + geom_polygon(data=map.df,aes(long,lat,group=group,fill=cat_num), 
                      color="white", size=.5)
p <- p + theme_minimal()
p

Looping over the three years

for (y in c(2000,2005,2010)){
  plot_data <- map.df %>% filter(Year == 2010)

  p <- ggplot()
  p <- p + geom_polygon(data=map.df,aes(long,lat,group=group,fill=cat_man), 
                      color="white", size=.5)
  p <- p + theme_minimal()
  p <- p + labs(title = y)
  print(p)
}