Overview

This chapter provides an overview of the three main pages of the Urban Analyst platform:

• Summarise providing summary overviews for each city of key statistical properties and relationships with other cities;
• Compare comparing aggregate statistics for all cities;
• Map showing interactive maps for each city;
• Transform for transforming any chosen city to become more like other cities.

The Summarise page is the best place to start. The summary for any chosen city will indicate which properties of the other pages are most important. Each of the latter three pages also includes a pop-up "guided tour" explaining the key features. These tours will automatically start the first time a page is opened, or can be manually opened by clicking on the "Help" button on any of the main pages.

Summarise

The Summarise page provides an overview of the general statistical properties of any chosen UA city. Statistics for each chosen city are compared with those for all other cities, and textual summaries generated for all statistics which are significantly different. Summaries are provided not just for individual statistics, but for the strengths of relationship between all pairs of statistics.

The values described in this initial part of the Summarise page indicate which statistics might be particularly worth examining in both the Compare and the Map pages.

The Summarise page also includes a section describing the best "Target city" for each chosen city. As explained there, the best target city is the city which is best in all the ways that the chosen city is worse than average. This target city, and corresponding variables described there, indicate which cities and variables might be particularly worth examining in the Transform page.

Compare

The Compare page is the first and main page of Urban Analyst, enabling the properties of all UA cities to be compared. A pull-down panel enables each variable or "layer" to be selected. The page then displays a graphical representation of values of the chosen layer for all cities. As in all UA pages, lower values are generally better than higher values. The control panel includes an "Explain Layer" button which opens a text panel explaining details of the chosen variable.

Single and Paired Variables

The control panel of the compare page includes an option to select "paired" variables. The resultant graphs then display the strength of relationship between any chosen pair of variables. For example, choosing social index and the nature index will display the strength of relationship between access to natural spaces and social disadvantage. Both of these variables are measured such that low values are better than high values. A positive relationship between the two would then mean that lower social disadvantage is coupled with better access to natural spaces, while high social disadvantage is coupled with worse access to natural spaces. Conversely, a negative relationship would indicate that higher social disadvantage was coupled with better access to natural resources. Or, in the words brought up by clicking the "Explain Layer" button, "Low values indicate that good access to natural spaces is coupled with disadvantageous social conditions."

Map

The Map page shows interactive maps for each city, with values for all UA variables displayed in small polygons. These polygons are defined by city-specific assessments of spatial disadvantage. Berlin, for example, regularly measures a compound index of social disadvantage aggregated into XX polygons. The map for Berlin uses these polygons provided by the city to aggregate all measured variables. The variables are described in a subsequent chapter.

As in all UA measurements, lower values of all variables are generally better than higher values. Colour scales on all maps thus generally display lower values in brighter, yellow colours, while higher values are displayed in darker, blue or violet colours. The control panel includes an "Explain Layer" button which opens a text panel explaining details of the chosen variable.

Transform

The Transform page enables the properties of any chosen city to be transformed to reflect equivalent properties of some chosen "target" city. This page is best explained by an example. Looking at the Compare page for the "bike index" shows that Berlin has relatively poor bicycle infrastructure, while Paris is a very good city for cyclists. The Transform page can be used to visualise how Berlin could best transform its current bicycle infrastructure to have an overall distribution across the whole city equivalent to Paris. Conversely, Paris has poorer access to natural spaces than Berlin, and the page could also be used to examine how Paris could best transform its access to natural spaces so that it functioned more like Berlin.

Urban Analyst values displayed in the Map page are aggregated from generally hundreds of thousands of individual calculations at every street junction in each city. For the chosen variable, subsets of these individual data points are sampled from each city, and the statistical distribution for the chosen city is then transformed by changing each point by the smallest amount possible so that they reflect the distribution in the target city. These values are then aggregated into the polygons defined for the city, to produce a visual representation of the least-cost transformation that would be necessary for the city to have the same distribution as that of the target city. The transformation algorithm is described in detail in the final Software and Algorithms chapter.

Extra Layers

The Transform page includes an additional button labelled Extra Layers. The transformations described above described transforming single layers or variables. The Extra Layers panel enables transformations not just of single chosen variables, but also of their relationships with other variables. Examining the Compare page, for example, shows that not only does Paris provide poorer access to natural spaces than Berlin, but also that Berlin has a better relationship between access to natural spaces and social disadvantage. (This can be seen by clicking on the "Paired" layer option and selecting those two layers.) The Extra Layers panel can be used in this case to examine not just how Paris might best transform its access to nature to look more like Berlin, but also how it might also improve its relationship between access to nature and social disadvantage.

By default, values of Extra Layers are automatically selected as those which have better relationships in the target city. These default values will thus change for each choice of target city and focal layer. It may be necessary to click on the "Reset" button in the Extra Layers panel to update this default selection after changing any of these options.

Output Layer

Finally, the Transform page also has an Output Layer option at the bottom of the control panel. This enables results of the transformation algorithm to be displayed in one of four ways:

1. Original to show original values, prior to transformation;
2. Transformed to show the actual transformed values;
3. Absolute to show the absolute value by which each are in the city would have to be transformed to match the distribution in the target city; and
4. Relative, which displays the absolute transformation values relative to the original, untransformed values.