无法在 R 中创建 India choropleth

Question

I want to create India's choropleth in R

The first step I do is import a shape file in R

From https://github.com/datameet/maps/tree/master/States

and read it in R

shape <- rgdal::readOGR(dsn="/Data/Admin2.shp")
states <- fortify(shape, region = "ST_NM")

Next I have a dataset of states and their population states_data

structure(list(Name = c("JAMMU & KASHMIR", "HIMACHAL PRADESH", 
"UTTARAKHAND", "RAJASTHAN", "UTTAR PRADESH", "BIHAR", "SIKKIM", 
"ARUNACHAL PRADESH", "NAGALAND", "MANIPUR", "MIZORAM", "TRIPURA", 
"MEGHALAYA", "ASSAM", "WEST BENGAL", "JHARKHAND", "ODISHA", "CHHATTISGARH", 
"MADHYA PRADESH", "GUJARAT", "DAMAN & DIU", "DADRA & NAGAR HAVELI", 
"MAHARASHTRA", "ANDHRA PRADESH", "KARNATAKA", "GOA", "LAKSHADWEEP", 
"KERALA", "TAMIL NADU", "ANDAMAN & NICOBAR ISLANDS"), TOT_P = c(1493299, 
392126, 291903, 9238534, 1134273, 1336573, 206360, 951821, 1710973, 
1167422, 1036115, 1166813, 2555861, 3884371, 5296953, 8645042, 
9590756, 7822902, 15316784, 8917174, 15363, 178564, 10510213, 
5918073, 4248987, 149275, 61120, 484839, 794697, 28530)), row.names = c(NA, 
-30L), class = c("tbl_df", "tbl", "data.frame"))

I merge both dataset on state names

final_data <- merge(states,states_data, by.y="Name", by.x="id")

Finally I plot using ggplot

ggplot()+
  geom_polygon(data=final_data,
               aes(x= long, y=lat, group=id, fill=TOT_P), color='black',size=0.25)+
  coord_map()

I get the following graph

Can someone tell me where I am going wrong. Any help is appreciated!

Thanks!

Answer 1

The strings of the state names are not identical across your two datasets.

If you take a look at the unique values, you can see that the shapefile uses title case

> unique(states$id)

[1] "Andaman & Nicobar Island" "Andhra Pradesh"           "Arunanchal Pradesh"       "Assam"                   
[5] "Bihar"                    "Chandigarh"               "Chhattisgarh"             "Dadara & Nagar Havelli"  
[9] "Daman & Diu"              "Goa"                      "Gujarat"                  "Haryana"                 
[13] "Himachal Pradesh"         "Jammu & Kashmir"          "Jharkhand"                "Karnataka"               
[17] "Kerala"                   "Lakshadweep"              "Madhya Pradesh"           "Maharashtra"             
[21] "Manipur"                  "Meghalaya"                "Mizoram"                  "Nagaland"                
[25] "NCT of Delhi"             "Odisha"                   "Puducherry"               "Punjab"                  
[29] "Rajasthan"                "Sikkim"                   "Tamil Nadu"               "Telangana"               
[33] "Tripura"                  "Uttar Pradesh"            "Uttarakhand"              "West Bengal"

while your population data frame uses all caps:

> unique(states_data$Name)
[1] "JAMMU & KASHMIR"           "HIMACHAL PRADESH"          "UTTARAKHAND"               "RAJASTHAN"                
[5] "UTTAR PRADESH"             "BIHAR"                     "SIKKIM"                    "ARUNACHAL PRADESH"        
[9] "NAGALAND"                  "MANIPUR"                   "MIZORAM"                   "TRIPURA"                  
[13] "MEGHALAYA"                 "ASSAM"                     "WEST BENGAL"               "JHARKHAND"                
[17] "ODISHA"                    "CHHATTISGARH"              "MADHYA PRADESH"            "GUJARAT"                  
[21] "DAMAN & DIU"               "DADRA & NAGAR HAVELI"      "MAHARASHTRA"               "ANDHRA PRADESH"           
[25] "KARNATAKA"                 "GOA"                       "LAKSHADWEEP"               "KERALA"                   
[29] "TAMIL NADU"                "ANDAMAN & NICOBAR ISLANDS"

That's why your merged dataset final_data is empty.

One possible fix is to turn the names in both datasets into lower case before merging:

states$id <- stringr::str_to_lower(states$id)
states_data$Name <- stringr::str_to_lower(states_data$Name)

However, there are still a few rows that will not be matched, either because of typos/different spellings or simply missing data. You could take a look at those via

setdiff(unique(states$id), unique(states_data$Name))

and where possible adapt the spelling.

Lastly, in my quick test the fortified polygons did not plot nicely -- this may entirely be specific to my combination of rgeos/rgdal/ggplot2. Still, in case you intend to work with spatial data more extensively, I would like to point you to the sf package. It makes handling spatial data extremely convenient (see the comprehensive documentation here ) and enables you to simply use geom_sf() for plotting with ggplot2 .

library(tidyverse)
library(sf)
# read shape and convert state names to lower case 
states <- st_read("./Data/Admin2.shp") %>%
                 mutate(Name = str_to_lower(ST_NM))
# merge spatial data with population data, also convert state names to lower case in the latter
states_population <- states %>%
  left_join(states_data %>% mutate(Name = str_to_lower(Name)), "Name")
# grey states are the result of unmatched states outlined above
ggplot(states_population, aes(fill = TOT_P)) +
  geom_sf() +
  scale_fill_viridis_c() +
  ggthemes::theme_map()