Popular Routes
Thanks for that additional clarification Mark. Much appreciated.
I totally get and accept that explanation and we can close it there.
Jim
Jim,
It seems to me that this is just the result of how the routing algorthim works. You select teh start and then the 1st point, and it calculates a route from the start to that point. When you click the next point, it does not want/need to recalculate teh entire route from the start, so it sets a new "start" as the finish if the first (last) section and the end as the last point, and auto-routes that section. In an ideal world, it would remove the extra point, but the process might incur some penalties that are just not worth the effort.
When using the auto-router point by point, what we may think of as a single resultant route is just a combination of many individual sections.
All the apps/devices I’ve used seem to manage this with no issues.
Regards
Mark
You are correct on all counts Mark. And it doesn't impact on my gps nor Fenix in the field. So from a practical perspective it has no navigational impact
And now that I have at least two viable workarounds to remove the duplicates I can still process the gpx data to produce my custom routecard.
So yes, a curiousity that Admin might want to look at sometime in the interests of consistency of outcomes for the product. That is, plotting a route by progressively extending it from the start point to the eventual finish point produces a different gpx download from one arrived at by reshaping an autoplotted route between the start point and the finish point - with the latter producing the'cleaner' gpx file.
So thanks to both you and Willy for your interest and insights and I guess we can leave it at that
Yes, Jim, it would have me about 20 clicks to copy the route.
In practice, the presence of these duplicate points does not seem to affect the usefulness of a plotted route. I have cycled thousands of miles following routes I created in PR with no issues at all.
Does it causes you a problem, Jim, or is it something you noticed and are seeking to understand?
So a question for Admin, then; is this a bug albeit one with (maybe) no visible impact?
On the ferry to Dublin, as I type!
Willy - thanks for those links.
I have found that editor and used it to remove duplicates.
Yet another viable workaround.
@ Jim.
You can analyse and solve that very quickle by the freeware gpx editor.
In the left down panel you find the anaylyse window. (Copy paste text)
(See also the you tube video) https://t.ly/-2M0
This issue anyway is caused by the GH engine.
A design by GH online.
https://t.ly/7252
Tap gpx for download.
Opening GraphHopper.gpx
6 waypoints.
0 routes (0 points)
1 tracks (55 points in 1 track segments)
TGPX
Waypoint list
Track Segment #1 (55 points)
3.6 km / 0:14:31 = 14.9 km/h
Remove duplicates.
Track Segment #1 (51 points)
Thanks for your detailed analysis Willy.
Your counts, give or take a point, mirror mine and appear to confirm duplicates in one file and none in the other.
Was that a manual exercise on your part (incliuding the removal of the duplicates)?
Mark - have a great trip around Ireland - my part of the world (obviously).
Yes, the routes will appear identical when shown on the map but if you look at the gpx file in a text editor you can locate the duplicates by checking the location co-ordinates of the trkpts. I also upload them to basecamp where they are easier to spot visually.
I downloaded your traced route and scanned the gpx data - there are in excess of 20 duplicates there. I suspect the number will reflect the number of times you clicked in tracing the route.
Opening Wr160b_11pts.gpx
1 waypoints
0 routes (0 points))
1 tracks (307 points in 1 track segments)
Track Segment #1 (307 points)
14.7 km / 3:59:29 = 3.7 km/h
Remove duplicates:
Track Segment #1 (295 points)
#
Opening Wr160c_Aps_Only.gpx
1 tracks (297 points in 1 track segments)
Track Segment #1 (297 points)
14.7 km / 3:59:30 = 3.7 km/h
Opening Wr160Traced.gpx
1 tracks (339 points in 1 track segments)
Track Segment #1 (339 points)
14.7 km / 2:56:18 = 5.0 km/h
Track Segment #1 (312 points)
I was a bit rash in offering to investigate, as I'm off on a cycling holiday to Ireland tomorrow. I looked at those 2 files are they appear to be identical, ie overlap when one is traced over the other, and are quite normal routes with nothing untoward that I can see.I am a bit confused when you talk of overlapping points; what app do you use to investigate the downloaded files?I downloaded wr16011pts as a gpx, then uploaded & created a trace. I then created a route using auto-route/On Foot/Streetmap using teh point and click method. See https://www.plotaroute.com/route/1947638See if that still gives you "duplicated points".Well, not a lot added but if still unresolved, I'll be back in 10 days or so!Regards,Mark
I can do that Mark.
Just made the two routes public.
They begin with wr160......
In answer to your questions - yes I used 'auto plot' and 'on foot' on Streetmap in both cases.
The route ending in '11pts ' was created using 11 points to develop the route from the start point to the finish point.
You will find when downloaded as a track that the gpx will have 11 duplicated points.
They should be at the follwoing track points....
9,10 34,35 65,66 77,78 136,137 146,147 192,193 228,229 244,245 253,254 294,295
The start and finish points are also duplicated.
The route ending in 'aps only ' was created using anchor points to reshape the initially autoplotted route between the start point and finish point.
Hope that illustrates the issue as I see it.
The route end in 'aps only' was created by using only a start and finish point and then reshaping it to be the same as the earlier one.
You should not find any duplicate points in that downloaded track.
If you could share your route, it would be an interesting experiment for me to trace it using the "standard" auto-routing method (which I have been using for years without the issue you have described) and see what happens. If so, what auto-route do ypou use ("on foot", "by bike"?) and what map type.
Hi Admin - I fully understand your position.
And while happy to close the issue now that I have a couple of workarounds I would suggest that two identical routes which when downloaded produce two very different gpx files should warrant a presence on the maintenance backlog. But then I would say that wouldn't I🙂.
Thanks again
New try....by link.
Remove duplicate trackpoints
Rather than using the Reshape Route tool, you could try the Snap Route to Map tool to replot it - this would be quicker and easier. We don't really see superfluous points on the route as a bug, as they don't affect the path the route takes, so this probably isn't something we'd be be able to look into given other priorities.
Thanks all for your comments.
Willy - can't access that youtube link.
admin - yes, I tried that reduction of points. It's a bit 'hit and miss' in that I progessively reduced the number of points until the problem disappeared but then the route wasn't quite mapping to the paths.
I also had an alternative solution - pre-processing the file to remove the duplications and that too worked and could be automated.
But then your observation about not knowing how I plotted the route triggered something and I suspected (and later confirmed) that all the duplications were at the points I used in guiding the plotting of the route. I had used the route plotting tool - selecetd the start point and then progresssivelly plotted the route to the destination point by left clickimg at the desired points along the way while allowing Plotaroute to map it to the available paths/roads.
So, I tried an alternative plotting method.
I left clicked on the start point and the finish point and let Plotaroute select a route between them based on a 'walking' profile.
I then used the reshaping tool to adjust that route to go where I actually wanted to go.
The outcome was a route identical to the previous one.
With one significant difference - no duplicated points.
Since this latter approach avoids the issue I can work with that.
That said, I must admit my original plotting approach is {to me} a more 'intuitive' method to develop a route from scratch and it would be great if it could be revised such that the duplications did not occur.
Finally, if this is indeed a bug in the system, how best to get it addressed?
Thanks again.
Or...
https://studio.youtube.com/video/NTn6ZHSpfR0/edit?o=U
It's hard to know without seeing the route and also knowing exactly how it was plotted. However, you can easily remove redundant points using the Reduce Points tool in the route planner (available on the route planner toolbar).
When I created and then downloaded a route in gpx format as a track (also true of tcx/fit formats) I noticed that a number of points had been duplicated - identical in every respect.
It usually happened at turning points in the route though not at all turning points and not only at turning points. But very much mostly so.
Is this a bug or an intentional feature?
And if the latter, why?
I won't bore you with the details but this duplication interferes with my currnet processing of the downloaded data to generate a custom route card.