uc_claudia

Polylines: Unsplit / Dissolve / Split line at vertices

Discussion created by uc_claudia on Mar 25, 2012
Hi

I am new to GIS.
I was looking for an answer in the forum, have found some posts to the same topic, but did not find the answer to my problem.


My original dataset has 24706 polylines.
Some of the polylines (not only two but sometimes 3 or more) have same endpoints.
Some are connected comparable to a T-junction (but with quite different angles, e.g. 10°)
Some of the polylines are not connected at all.

There is no rule at the moment and that's why I would like to transform the data so that I can describe the polylines.

I tried different ways and got quite different results. I would like to understand the reason for these different results.



part 1: using 24706 polylines as input feature

- unsplit ->                                                           I get    11635 polylines
- dissolve (single part, unsplit) ->                                        11635 polylines (same number makes sense)
- dissolve (single part, dissolve) ->                                      12044 polylines (should be less than 'unsplit' as 'unsplit' only dissolves if polylines have the same endpoint)

Question 1: What criteria uses 'dissolve' to dissolve? (e.g. for 'unsplit' it is the same endpoint)



part 2: splitted the 24706 polylines resulting in 267664 polylines and used these 267664 polylines as input feature

- unsplit ->                                                           I get    11834 polylines
- dissolve (single part, unsplit) ->                                        11834 polylines (same number makes sense)
- dissolve (single part, dissolve) ->                                      11694 polylines (less polylines make sense as 'unsplit' only dissolves if there is an endpoint)

Question 2: I thought that I should get the same numbers as for part 1. Why not?



part 3: dissolved (multipart, dissolve) the 24706 polylines resulting in 1 polyline and used this single (but big) polyline as input feature

- splitted at vertices ->                                      I get    127973 polylines

In my case this tool seems to be useless as I even got more polylines than my original data set was.



Question 3:
After doing this and getting all these different results, I think I need a tool with which
- I can unsplit lines if they have the same end point but only if the angle between them is 180° +/- 10° (170°-190°).
- I can find the points at which 2 or more polylines end (but do not have an angle between 170°-190° and therefore are not unsplitted).
- I can find the points at which are connected comparable to a T-junction (even the angle is quite different, e.g.10°)



Question 4:
Can you recommend any literature for analyzing line networks? Most of the literature I found are only for point, area and field data.



Thank you very much in advance for your help

Claudia

Outcomes