2

I am using ArcGIS Desktop 10.2.

I have created the origins points from a fishnet laid over my area of interest. Out of 1200, there are some 65 points for which the OD cost matrix cannot be calculated.

I have isolated these 65 points and tried to run the OD cost matrix only for them (just in case they were not computed because the matrix was too big), but to no avail. The warning is still displayed for all 65 locations. Eg: for location 1 "No destinations found for "location 1" in "origins"".

Any ideas on how to solve this problem?

Legend of Picture: yellow dots are those O that were successfully calculated, the orange are those that were not successfully calculated and the violet lines are each OD cost link.

enter image description here

1
  • What are your origins and what are your destinations? 1200 origins? 1200 destinations? Explain your OD Cost Matrix analysis settings. Commented May 8, 2017 at 19:08

2 Answers 2

1

The problem is not so much the OD Cost Matrix tool, but the network dataset you are using.

The path between your origins and destinations is done along a road network (based on a network dataset) and those 65 that could not be located means that between those 65 and the locations they are going to/from, the OD Cost Matrix tool could not find a way to get to those 65 with the network dataset you're using.

What you need to do is take another look at how your points were located into the network (search tolerance? network locations?) and if there are any restrictions you have turned on that are preventing the OD Cost Matrix tool from getting to those 65. Most likely that's the culprit.

I've had this happen to me using the ESRI StreetMap Premium network dataset. When I ran the OD Cost Matrix, 3 of my destinations could not be located and it was due to the default "Driving an Automobile" Restriction in the OD Cost Matrix properties window.

0

Excuse my late reply!

The problem laid with the dataset itself not with the tool. Thanks a lot for your help. Really appreciate it.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.