Hello again,
I'm testing Routing 6.0 and we've noticed from time to time when the routing result is drawn on the map there are gaps. The boss isn't terribly pleased with that. Below are a two examples.
Is this a known problem? Anyone else see this? I haven't gotten brave enough to actually look at the feature that's being returned to see if there actually is a gap in it or something else is going wrong.
Allen
I set a breakpoint in Visual Studio to capture the WKT of the resulting route and then pasted it into Management Studio to see it better. I zoomed in and there are most definitely some gaps in the route that has been returned...in fact, two of them with a little sliver in between. Since nobody has responded to this, I'll probably have to make it a support ticket.
Here is the WKT…
declare @g geometry
set @g = ‘MULTILINESTRING((-8494781.1935 4842686.6099,-8494754.0522 4842449.4198,-8494720.5558 4842236.3448),(-8494788.929 4842748.6011,-8494781.1935 4842686.6099),(-8494810.9709 4842908.5046,-8494788.929 4842748.6011),(-8494819.5627 4842966.7715,-8494810.9709 4842908.5046),(-8494870.7844 4843043.9788,-8494826.8803 4843005.6546,-8494819.5627 4842966.7715),(-8494973.5475 4843130.102,-8494870.7844 4843043.9788),(-8494985.7166 4843139.6807,-8494973.5475 4843130.102),(-8495015.927 4843164.4626,-8494985.7166 4843139.6807),(-8495050.0501 4843193.1288,-8495015.927 4843164.4626),(-8495083.487 4843221.2246,-8495050.0501 4843193.1288),(-8495117.1175 4843249.6465,-8495083.487 4843221.2246),(-8495179.8787 4843302.0747,-8495117.1175 4843249.6465),(-8495254.8986 4843358.6408,-8495245.7706 4843357.6164,-8495179.8787 4843302.0747),(-8495331.2101 4843409.1829,-8495269.9818 4843360.1968,-8495254.8986 4843358.6408),(-8495406.9492 4843470.5312,-8495331.2101 4843409.1829),(-8495395.5909 4843484.4179,-8495406.9492 4843470.5312),(-8495442.1214 4843655.0673,-8495405.1943 4843505.131,-8495395.5909 4843484.4179),(-8495449.2026 4843684.2136,-8495442.1214 4843655.0673),(-8495464.1227 4843744.8114,-8495449.2026 4843684.2136),(-8495479.5195 4843806.6784,-8495464.1227 4843744.8114),(-8495491.5934 4843865.0561,-8495479.5195 4843806.6784),(-8495491.5934 4843865.0561,-8495523.9845 4843878.1876),(-8495523.9845 4843878.1876,-8495574.2411 4843892.9943,-8495695.6118 4843935.9608),(-8495695.6118 4843935.9608,-8495740.7186 4843949.6147,-8495878.4536 4843964.2906,-8495909.2747 4843976.2026),(-8495909.2747 4843976.2026,-8495945.9122 4843972.7016,-8496071.7668 4843984.4365),(-8496071.7668 4843984.4365,-8496109.8759 4843988.2774),(-8496109.8759 4843988.2774,-8496284.737 4844006.1823),(-8496284.737 4844006.1823,-8496298.9333 4844007.3311),(-8496298.9333 4844007.3311,-8496444.7975 4844024.3993),(-8496444.7975 4844024.3993,-8496447.0644 4844024.6646),(-8496447.0644 4844024.6646,-8496455.6964 4844025.6675),(-8496455.6964 4844025.6675,-8496552.0592 4844035.5348),(-8496552.0592 4844035.5348,-8496600.8837 4844040.6587),(-8496600.8837 4844040.6587,-8496752.1323 4844055.8199),(-8496752.1323 4844055.8199,-8496899.7941 4844071.5176),(-8497053.9305 4844085.9799,-8497061.3817 4844086.679),(-8497052.7497 4844118.437,-8497003.0592 4844301.2539),(-8497003.0592 4844301.2539,-8497398.9264 4844386.306))’
select @g as geom
I’ll snip out just the last five or six segments so that their coordinates can be inspected:
(-8496600.8837 4844040.6587,-8496752.1323 4844055.8199) <-- This one touches the next segment
(-8496752.1323 4844055.8199,-8496899.7941 4844071.5176) <-- This one does not touch the next segment
(-8497053.9305 4844085.9799,-8497061.3817 4844086.679) <-- This segment doesn’t touch the previous or next one (must be the little piece in the gap)
(-8497052.7497 4844118.437,-8497003.0592 4844301.2539) <–This segment touches the last (next) segment
(-8497003.0592 4844301.2539,-8497398.9264 4844386.306) <-- Last point in route
So there definitely are two gaps in the line shape being returned.
Hello Allen,
Thanks for your post and sorry for delay, you did really deep research, thanks for your hard work, from the result wkt and your analysis, looks like the error reason is routing matrix calculate wrong route segments, could you please provide your test sample and data, then we can debug and fix it?
Regards,
Gary
Hello Allen,
Thanks for your post and sorry for delay, you did really deep research, thanks for your hard work, from the result wkt and your analysis, looks like the error reason is routing matrix calculate wrong route segments, could you please provide your test sample and data, then we can debug and fix it?
Regards,
Gary
Gary,
This is our lucky day…the data for this forum item is the same as for my posting under "Unexpected route" where the route would zigzag in a weird manner. I used Routing Explorer to generate the .rtg file, so I think all I need to send to you is the original shapefile. Sound reasonable?
Allen
Gary, I just uploaded "roads.shp.zip" to the FTP site…our user name is "ironcompass" and this file is under the "Routing Data" folder.
I fed this shapefile to Routing Explorer to generate the .rtg file and then used code copied from an example to plot the route on a map. Here are three to/from ID pairs that show a gap in the route:
116214 -> 122446
143049 -> 144996
118990 -> 144996
I also have one route that does the unusual zig-zag, but I will post that in the other posting about that issue.
Allen
Gary,
I had not tried this before, but I get the gaps when using Route Explorer. Here's a closeup and a zoomed-out image so you can see where in the county the route is...these may not be the exact road segments because I couldn't get Route Explorer to label the roads, but there is still a gap. (This is the city of Lancaster, PA, which is almost in the middle of the county...to help you get your bearings if you have trouble finding where the overview map is located.)
Allen
Sorry, "Routing Explorer" not "Route Explorer".
Hello Allen,
Thanks for your data, I got it, after I did the test, I will let you know the result and hopeful we will find the root cause and fix it very soon.
Sorry for the inconvenience,
Regards,
Gary