This section outlines the iterative process of development we used to produce Priestley’s charts in a data driven, procedural, way. We processed the chart data using desktop GIS tools, then display the graphics and related charts using web-visualization tools.
At its core, this spreadsheet is the data behind the digital chart. Each cell represents which regimes were in power for each year, and codes are used to represent the transitions in power. There are around 140 rows (places) and 3000 columns (years).
The highlighted are represents the group Heruli, who ruled Italy from 475 - 489 C.E. The orange box above hihglights Vandals, showing the structure of the spreadsheet.
This image shows the same highlighed area from the spreadsheet in gis (left) and in the map (right). THe contiguous cells withthe same values are merged into polygons that represent the total amount (area, time) they ruled an area.
From the spreadsheet (CSV) the values are translated into ASCII (names to numbers). The ASCII file can be read by Esri ArcMap to become a graphical representation.
Using Python and ArcPy, the different transitions are created with the codes that determine whether the split is diagonal or vertical. The split polygons are then dissolved into the corresponding regions. This is then saved and exported to JSON, which is the format of the chart data used on the site.
The chart data is stored as distinct layers (regions, places, regimes, event lines, etc). In this image, distinct symbology emphasizes the different layer polygons.
Here you can see the finished product. The digital, procedurally created one looks a lot like the original. How neat is that?!