Spotfire Tips & Tricks: Create Better Maps using Tiles
Last updated:
3:31pm Apr 24, 2018

Overview

Choropleth or thematic maps are the default way to visualize geographical information. These maps display divided geographical areas or regions that are coloured, shaded or patterned in relation to a data variable[1].

Fig. shows a Choropleth State Map in Spotfire

But, choropleth maps have a number of known shortcomings in encoding color information. Let’s say you use variable ‘X’ as the color by attribute on your map chart, this could lead to perceptual distortions in judging the importance of certain regions.

Perceived importance of a region is the color value of the region multiplied by the area of the region[2]. So small regions with high color intensity can often seem less important than large regions with high color intensity despite equal values of variable ‘X’.


Fig. shows a Grid  State Map in Spotfire

This can be overcome by using Tile maps. In Tile Grid Maps, the distinct geographic regions are reduced to a single uniform shape of constant size. Once the area is held constant, our perceived importance equals the aggregation measure for variable ‘X’ on that state.

Create tile maps in Spotfire

Note: Completed Tile Map DXP is attached, but the following steps will help you use the same idea to tessellate world map, maps of other countries, part to whole relationships etc.

1. Choose a reference map

Tile maps exist with different shaped tiles as well as different spatial approximations to state positions [3]. Choose a reference map based on these characteristics or see how you can create your own using R [4].

Fig. Shows square tile grid map chosen as reference. Source.

2. Create axis points

To encode the position information of the states, we’ll mark a grid of x,y coordinates on the map as shown below. Note the Center position for each state, for example, Alaska is centered at 1,15. Origin or 0,0 is the left bottom corner.

3. Create state position data

Using the grid I can create a dataset of State name and location. This is attached as csv. For convenience I have included state name as well as state abbreviation codes

4. Create scatter plot

Import this information in Spotfire and create a scatter plot. We can see the grid tile map taking shape.

 

5. Additional configurations

  • In scatter plot Properties > Shape > Shape definition, choose square

     
  • In scatter plot Properties > Size, increase marker size until tile effect is created


    Fig. Shows the tile effect in scatterplot
     
  • In scatter plot Properties > Size, show labels for marked

     
  • Hide X, Y axis selectors
  • In scatter plot Properties > Y-Axis, unselect Show Labels. Same for X-Axis

     

6. Connect your data

You can connect your data to the tile map visualization by:

  • Inserting X,Y columns in your existing data set, matched by State name or abbreviation and doing the above steps
  • Inserting the column from another data set into the current using various spotfire join operations

Fig. Shows complete State Tile Map visualisation in Spotfire
Instead of Color by, you could also use your data in Size by and display at Marker tile map.

How do I learn more?

This completes the tutorial on creating tile maps in Spotfire. Watch the page and vote up to get notified about detailed updates. You could also request a featured session on any specific method from above on Dr. Spotfire by:

Citations

  • [1] Choropleth Maps - Wikipedia. (2017). Retrieved from URL
  • [2] G. Aisch, Take Care of your Choropleth Maps. Retrieved from URL
  • [3] N. Yau, Great Grid Map Debate, published on FlowingData (2015). Retrieved from URL
  • [4] Nick, Visualizing Geographic Data in R, published on Medium. Retrieved from URL
     

See Also

  • (Wiki) Spotfire Tips & Tricks
  • (Wiki) Spotfire Location Analytics: Mapping, GeoAnalytics & Spatial Analytics
  • (Reference) Let’s Tessellate - Hexagons for Tile Grid Maps
  • (Reference) Spotfire Insert columns

Attachments

AttachmentSize
Plain text icon statetile.txt951 bytes
Binary Data tile_state_map.dxp709.11 KB