Email discussion lists for the UK Education and Research communities

SPACESYNTAX@JISCMAIL.AC.UK

View:

 Message: [ First | Previous | Next | Last ] By Topic: [ First | Previous | Next | Last ] By Author: [ First | Previous | Next | Last ] Font: Proportional Font

Options

Subject:

From:

Date:

Thu, 22 Mar 2018 08:20:35 +0100

Content-Type:

text/plain

Parts/Attachments:

 text/plain (51 lines)
 ```Dear José, That is a complex question that could also be placed on the general space syntax mailing list, as here we try to address aspects of the QGIS toolkit in particular. Nevertheless, I'll try a simple answer... Global analysis is always radius N: to everywhere. Therefore its results are the same whatever radius type you use, metric, angular or steps. Metric type radius can be any value, and there are studies that use radius from 100m up. Keep in mind that long segments with more than 2x the radius don't get any integration value for that radius because you can't get anywhere from them. The choice of radius must be linked to the research problem and hypotesis being tested. If your problem relates to pedestrian movement, 400m makes sense as it corresponds to 5 minutes walk, and in much walkability literature is regarded as a threshold after which the propensity to walk drops. And if your main focus is the car, then larger radius make sense. But you might be looking at centrality structures irrespective of behaviour, then the choice of radius must come from some hypothesis. In many studies you find very large radius (60km) just because N would be too slow to calculate in large models. Steps radius is rarely used in segment analysis. That would be limiting the range of centrality based on the number of segments (maybe axial lines, not sure). This is the standard in axial maps as your theory considers changes of direction. In principle it can be used as long as your hypothesis is that centrality matters up to a certain number of steps. Angular type radius is even more rare, and the same principle apply: you must have a theory or hypothesis to use it. It's also more difficult to define and understand, as it sets the limit of degrees that you can turn. These sorts of radii are very irregular, allowing very long distances (many km) along street alignments, and not going far into adjacent grids. Hope this clarifies things a bit, but as I said, try also the main list for further insights. Best regards, Jorge On 22/03/2018 06:44, Jos é Rojas wrote: > Hi all, > checking the tutorials on the SS Online Platform I saw that examples use metric radius type and I understand that it could be from 400 to N meters but I want to know if there is a recommended radius for the angular analysis and segment step analysis too. From which radio can a global analysis be differentiated from a local analysis? > I will appreciate any help.```