1.0.5 NULL Values for "OD Matrix from Layers as Lines"

[jrbarnard01]

Hello @root676, hoping you could assist me with this. I am trying to use "OD Matrix from Layers as lines" tool to create a shapefile of the fastest optimal routes between each mexican capital.

When I run the tool with red_vial (EPSG 6365 - shape file of road system across mexico) as the network layer and the points of each state capital as the start and destination locations, I get a result that has one or two lines displayed with the optimal route, but the rest is NULL. This has worked in the past with the same dataset, so it must be something weird. Could you help me with this? I am using 1.0.5 version of QNEAT but I don't see where I can adjust the tolerance (as mentioned in a previous thread). Here is what my results look like...

[root676]

Hi!
The codebase of QNEAT3 did not change since january 2023. Did you use another version of the road network? This usually happens with mismatching coordinate systems of points and the graph. Other solutions might be to use the builtin QGIS network analysis algorithms (from the processing toolbox) to test if the routing is possible. QNEAT3 and QGIS use the same network analysis library in the background.

[jrbarnard01]

I reprojected my red_vial layer to EPSG: 4326 and had my point layer to also be EPSG: 4326 and it still didn't not work as intended. Any thoughts as to why this is the case? Interestingly, it worked a couple days ago, but not now. I don't recall changing anything. I will test the "bulletin QGIS network".

[jrbarnard01]

@root676 The issue is with the topology: when I increase the topology, while there a more values that are calculated, the routes are les precise. How can I adjust the settings so it incorporates every state in the calculation while remaining precise in following the red_vial route? Thank you for your time.

[root676]

If that is so, then you have an issue with your graph topology. I suggest checking connectivity between graph edges.

[jrbarnard01]

Would I check the connectivity of the graph edges for the network layer?

[jrbarnard01]

@root676 You are definitely right, I did some more testing and got the same result. However, what's interesting that I did this a few days ago with no systematic changes and it worked perfectly.

I wonder what could have changed... Do you have any suggestions of how I can fix this?
Again, I really appreciate you taking the time to look at this. It means a lot.