Our aim is to promote the practice of spatial network analysis through
- Provision of training and accreditation
- Provision of software
- Provision of a knowledge base as a foundation for evidence based decisions and policy.
The sDNA methodology differs from previous forms of spatial network analysis by standardizing on the network link as a unit of computation. All Euclidean spatial analysis is subject to the modifiable areal unit problem, which makes comparison of results from different data and different areas problematic. However, networks are fundamentally different to Euclidean space not least because they are composed of atomic, indivisible units of a sensible scale for everyday use: network links.
By standardizing on the network link, our methodology mitigates problems caused by varying spatial scales, which allows results from one study area to be transferable to another (hence allowing us to collect an evidence base for adoption in guiding best practice in network design). Our continuous space enhancement to existing network algorithms also allows us to ensure that accuracy is not lost by treating links as atomic rather than divisible entities.
Finally, our methodology is easy to use with existing map data, and we support two workflows: Autocad for architects and designers, and ArcGIS for analysts and researchers. We plan to support more workflows in future, making results directly comparable over a wide range of software platforms.
The sDNA Logo is actually a spatial network coloured according to one of the sDNA Output Measures.
Can you guess which one?
(Hint: the representation is two dimensional apart from the links crossing over where the helix twists)