Network City I // Distances in a Graph – Representation problems

[Kevin Carvalho]

Hello, mr. Reinhard

First of all, congratulations for the work in DecodingSpaces, again.

Well, I’ve applied successfully the network calculation with the Betweeness Centrality Weighted results, and I exported the results to shapefile (via BearGIS, a nice plugin for shp exporting). Now, comparing the shapefile view in ArcMap (from ArcGIS) to the visualization in Rhino, I see a giant difference. In GIS, I see the line with the right colours, according to its values, while in Rhino I can’t get the same result.

See the images. In grasshopper (right image), gradient receives the “values” from 0 to 1, so it should show blue streets in lower values and red in the bigger values. It doesn’t happen in GH, that doesn’t show blue lines at all, while in ArcGIS (left image) it happens in the right way.

What is happening in Rhino that doesn’t show the same result?

The gh file is attached.

Thank you in advance,
Kevin

[Kevin Carvalho]

The attachment didn’t work, i’ll try again here, with the ZIP. Also, I add the attribute table’s printscreen from ArcGIS, so you can see the index numbers and how it matches with Rhino. 

Attachments:

You must be logged in to view attached files.

[Kevin Carvalho]

Well, sorry for the long post.

But I saw the problem. Rhino gives value 0.00 to the blue and 0.50 to the yellow and 1.00 to the red, while ArcGIS doesn’t give colours to numbers that doesn’t exist. For example, there is no 0.20 value: 0.20 would be blue in GH while 0.40 would be yellow. ArcGIS jumps the values that does not exist, so 0.40 is still blue. So, how can I make GH give colours the same way GIS does?

[Reinhard Koenig]

Dear Kevin,

you need to adjust the upper or lower bound value for the color gradient – add a value to it and you’ll see also blue lines in Grasshopper.

Concerning your last post, you can filter the invalid or 0 values.

Best regards

Reinhard

[Kevin Carvalho]

Thank you for the answer, mr. Koenig.

Also, I’m analysing another result. When the street network has avenues with double lines (one for each way), as you can see in the image bellow, I wonder what means the result: if it’s a medium centrality (yellow), if it was only one segment, probably it would be red? I’m wondering how the script deals with those kind of representation, double lines for only one avenue. I’ll do the test without that kind of street, but I would like to know what you think.

Best regards,

Kevin

[Reinhard Koenig]

Dear Kevin,

The issue with parallel lines for the “same” street is a problem for all kinds of network analysis. The only way to deal with it is to clean the data before the analysis (=merge the two street center lines). Also, the roundabouts in your map could distort your analysis if they consist of many small line segments (=many origins and destinations). There are tools for simplifying street networks (in QGIS, ArcGIS, and not that advanced in the DeCodingSpaces-Toolbox), or you do it manually.

Best regards

Reinhard

[Author]

Posts