I have polylines from a network analysis representing public transport trips. These are multipart polylines.

I need to identify the starting point for each trip. However, using vertices to points to create starting points, I get starting points for every part of the line. To get only one point, I tried transforming the lines with multipart to single part and then unsplit, as suggested here, as well as dissolve but none of them worked.

Fyi, the lines have IDs.

  • What exactly 'did not work'? Because it looks as though it should. If all else fails, you could add two extra columns: x and y, and use the field calculator to set those to !SHAPE!.firstPoint.x and !SHAPE!.firstPoint.y. Copy the values to a table and add it as an event layer. – Berend Nov 29 '19 at 15:45
  • @Berend: The "vertices to point" creates one point feature for every part of the multipart polyline. In that sense, the tool doesnt work for my case. So I tried splitting the multipart line into its parts to then dissolve it into one line with only one part in order to get only one point with 'vertices to points'. But that does not work either. With dissolve, I just get te old multipart polyline again, when I allow for multipart. If I don't allow for multipart, None of the features get dissolved. – jpg Nov 30 '19 at 21:02
  • @Berend: Your idea with the XY layer was the way to go. Thanks a lot! – jpg Dec 1 '19 at 0:15
  • I guess I have to ask the obvious question, if you generated the polylines in network analysis don't you have the starting locations already? If this is not the case, it will help others to understand what your data actually is if you edit your question and put up several screen shots of why vertices to pints is failing you. – Hornbydd Dec 1 '19 at 1:02
  • Thanks for your comment. You are right. I should have starting points (incidents). But I forgot to save them before loading new incidents into the network analysis to do another analysis. With your comment in mind, I reassessed the data and found out the routes generated were not even multipart. However my colleague merged and dissolved several feature classes (from analysises from different points in time) to get a mean travel time. And this new feature class was the problem (and still is). However retrieving X and Y from the original routes as suggested by @Berend was the solution – jpg Dec 1 '19 at 10:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.