A diverging scale example

Let's take a look at another dataset where the default color scale isn't appropriate. This raster, migration, has an estimate of the net number of people who have moved into each cell of the raster between the years of 1990 and 2000. A positive number indicates a net immigration, and a negative number an emigration. Take a look:

tm_shape(migration) +
  tm_raster() +
  tm_legend(outside = TRUE, 
            outside.position = c("bottom"))

The default color scale doesn't look very helpful, but tmap is actually doing something quite clever: it has automatically chosen a diverging color scale. A diverging scale is appropriate since large movements of people are large positive numbers or large (in magnitude) negative numbers. Zero (i.e. no net migration) is a natural midpoint.

tmap chooses a diverging scale when there are both positive and negative values in the mapped variable and chooses zero as the midpoint. This isn't always the right approach. Imagine you are mapping a relative change as percentages; 100% might be the most intuitive midpoint. If you need something different, the best way to proceed is to generate a diverging palette (with an odd number of steps, so there is a middle color) and specify the breaks yourself.

Let's see if you can get a more informative map by adding a diverging scale yourself.

(Data source: de Sherbinin, A., M. Levy, S. Adamo, K. MacManus, G. Yetman, V. Mara, L. Razafindrazay, B. Goodrich, T. Srebotnjak, C. Aichele, and L. Pistolesi. 2015. Global Estimated Net Migration Grids by Decade: 1970-2000. Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC). http://dx.doi.org/10.7927/H4319SVC Accessed 27 Sep 2016)