Visualizing international migration flows

1 | Shiny app pitch - introduction

A shiny app for visualizing international migration flows to and from a region using UN migrants stock data.
Colors to visualize direction of movement, and number of arcs to visualize size of movement.
It's very easy to use, just select the region, theme, year to update the output. That's it!

Try it at https://asheshwor.shinyapps.io/migrationviz/

Screenshot of options

@asheshwor | 2014-07-27

2 | Getting the data

The migration data was obtained from United Nations, Department of Economic and Social Affairs, Population Division
The world map shape file was obtained from NaturalEarthData.com. The location and population of cities were obtained from the cities database at geonames.org
Code for reading migration data, world map shape file and cities database:

data2013 <- read.xlsx2("data/UN_MigrantStockByOriginAndDestination_2013.xls",
                       sheetName = "Table 10",
                       startRow = 16, colIndex = c(2, 4 , 10:241),
                       colClasses = c("character", rep("numeric", 232)))
wmap <- readShapeSpatial("data/110m_cultural/ne_110m_admin_0_countries.shp")
places <- read.csv("data/cities1000.csv", header=FALSE, stringsAsFactors=FALSE)

3 | Data processing

With some processing, a data-frame with the required arc connections is created. Following is an example a section of the dataframe for Australia

##   source destination stock lat.d lon.d lat.s lon.s stocklog id
## 1     AD          AU    22   -27   133  42.5   1.5        3  1
## 2     AD          AU    22   -27   133  42.5   1.5        3  1
## 3     AD          AU    22   -27   133  42.5   1.5        3  1
## 4     AE          AU  5890   -27   133  24.0  54.0        9  2
## 5     AE          AU  5890   -27   133  24.0  54.0        9  2
## 6     AE          AU  5890   -27   133  24.0  54.0        9  2
## 7     AE          AU  5890   -27   133  24.0  54.0        9  2
## 8     AE          AU  5890   -27   133  24.0  54.0        9  2

In the next step, the source and destination coordinates are replaced with locations of cities from the country or region.

4 | Migrant origin and destination points

The coordinates for each origin or destination in a region are sampled from 15 most populated cities in that region with probability based on the population.
Since the sampling is done at random, the map generated each time is different.

More populated cities are more likely to get selected

Cities with higher population are more likely to get selected

5 | Generating the final map

Finally the great circle arcs obtained using gcIntermediate function are and plotted using ggplot2 over the world map.

plot of chunk final-plot

An example plot for Australia with 'light' map theme

Try the app at https://asheshwor.shinyapps.io/migrationviz/