update walkability index approach to work with an expanding set of cities

[carlhiggs]

The global indicators software tool was initially designed as a proof of concept spatial indicator analysis workflow for 25 cities. However, to work for an open ended number of cities, a walkability index cannot be calculated based on sub-components scored relative to the observed values across all cities. This would mean all cities walkability scores would require re-calculation as each new city were added. Instead, benchmarking against some fixed target(s) which are ideally policy relevant, could be desirable.

This would be relatively easy to implement, but determining which are the relevant thresholds that could apply as appropriate targets across different cultural, economic and geographic contexts

[carlhiggs]

I have proposed to others that we resolve this problem essentially by removing the between-city walkability index for the global project, at least until we develop and validate a revised method we all agree upon. (fyi @VuokkoH )

That is, we present 'walkability' not as a composite index, but rather as a collection of indicators for street connectivity, dwelling density and access to a range of services and amenities

• pros:
  • very do-able,
  • defers the 'how to do walkability problem' by presenting the core information, and allowing others (perhaps us) to address the 'how' later on;
  • contextually-relevant thresholds could be presented in prose
  • within-city walkability index could still be calculated, if desired
  • the more advanced methods could be incorporated at a later stage
• cons:
  • our current spatial maps and statistics for walkability relative to other cities can't be produced as part of the automated process, until a validated method is incorporated,
  • nor can we generate the percentage of population in high walkability neighbourhoods

Essentially, I think this is an easy way to side-step the composite index problem while retaining the core walkability-related measures which would allow scores and comparisons against policy- and contextually-relevant thresholds to be calculated at a later stage as needed, and even incorporated into generated reports via supplied prose.

Hopefully we come to a decision in the coming week, and then we can update the code accordingly!