Skip to content

Latest commit

 

History

History
113 lines (81 loc) · 3.58 KB

README.md

File metadata and controls

113 lines (81 loc) · 3.58 KB

Sample-Mapping-With-R

Step by step on how to map with R.

Prerequisite

Basic knowledge in GIS.

Installation

R and Rstudio
For this tuto, R version 4.2.2 Patched (2022-11-10 r83330) is used.

Package

R Package needed for this first tuto is rgdal which provides functions for reading and writing vector and raster spatial data.

FYI, here is the website for all spatial packages available : https://cran.r-project.org/web/views/Spatial.html

PS : In this tuto, we use the Madagascar 23 regions shapefile from this link : https://www.unsalb.org/data/mdg?fbclid=IwAR0bVCV5CJdmKy833ESaY9bbERWo6AbI9Jdv5EasQFpnAkcl_IYoRZ6NeMQ

Codes

First of all, load the rgdal package

library(rgdal)

Next, assign the new spatial object within a variable named mada_regions , and use the function readOGR in order to load the shapefile

mada_regions <- readOGR(dsn = "data/mada_23_regions.shp")

Check the structure of the data

str(mada_regions)

Check the attributes and their type

sapply(mada_regions@data, class)

Plot all regions

plot(mada_regions, col = "lightgrey")

mada_grey

Select region who has the largest area

In here, we use the Shape_Area column to check the value.

big_r <- mada_regions$Shape_Area == max(mada_regions$Shape_Area)

Add selected zones to map

plot(mada_regions[big_r, ], col = "orange", add = TRUE)

Add a title to the map

title("Madagascar, the 23 regions")

mada_with_highest_r

To conclude, this is just a simple manipulation to begin with. We will see for the next coming tuto, thematic maps with some spatial analysis!

If you would like to further your skills in mapping with R, there are lots of resources out there, not to mention https://github.com/Robinlovelace/Creating-maps-in-R

Join more attributes from external source

In here, we load a csv which contains the statistic data related to number of habitants in Madagascar that I get from the internet (means it is not an official dataset, just for the practice)

stat_pop <- read.csv("pop_region.csv",
                       stringsAsFactors = FALSE)

Now, we compare the ADM2NM column in mada_regions to region column in stat_pop to see which rows match.

mada_regions$ADM2NM %in% stat_pop$region

The result indicates that all names match

[1] TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE
[20] TRUE TRUE TRUE TRUE

Join the spatial and non-spatial datasets (the csv) by loading the dplyr package

library(dplyr)
mada_regions@data <- left_join(mada_regions@data, stat_pop, by = c('ADM2NM' = 'region'))

Plotting with aesthetic

In here, we use tmap package

library(tmap)  
tm_shape(mada_regions) + tm_polygons(col = "pop", palette="Reds", title="Number of Population", style="jenks") + tm_layout(bg.color = "lightblue")+tm_scale_bar(breaks = c(0, 100, 200), text.size = 0.7)
  • tm_shape takes the input shapefile.
  • tm_polygons speficies the column associated with the geometry to get mapped, the color, the title of the legend, and the style for the classification (we use jenks style here)
  • tm_layout for background
  • tm_scale_bar for the scale bar
  • You can check various alternative in here https://geocompr.robinlovelace.net/adv-map.html

image