[BUG]: Error found on redefine_old_data_to_current_time #76
Description
Version
25.02
Which installation method(s) does this occur on?
Docker
Describe the bug.
When reoptimizing, the returned routes are not correct. For the basic example given in the CuOpt NIM page, reoptiziming with two more pickup/delivery pairs can return routes of the type: [depot] ---> [pickup] ----> [delivery] with route numbers [32, 32, 32, ...].
This happens when doing the following edge case:
- You optimize once
- You set a time rerouting period of 700 minutes (from 00:00h)
- You have some pickups within some vehicles that have not yet been delivered.
- You have some pickups/delivery pairs completed
- The resulting routes stack the same location for the depot, the pickup and any other pickup / delivery sometimes, resulting in a route like the one mentioned above.
Minimum reproducible example
Using Nvidia exact examples csv from Pickup and Delivery: https://build.nvidia.com/nvidia/nvidia-cuopt?ncid=no-ncid
Adding 700 reroute minutes time and the following csv:
name,order_ID,service_time,order_wt,order_type,delivery_start,delivery_end,Latitude,Longitude
pickup_50,10001,30,40.5,Pickup,2022-03-18T09:00:00,2022-03-18T17:00:00,40.75000,-73.98000
pickup_51,10002,45,22.3,Pickup,2022-03-18T10:00:00,2022-03-18T18:00:00,40.75500,-73.97500
delivery_50,10003,40,-40.5,Delivery,2022-03-18T09:30:00,2022-03-18T17:30:00,40.75100,-73.98100
delivery_51,10004,50,-22.3,Delivery,2022-03-18T10:30:00,2022-03-18T18:30:00,40.75600,-73.97600
And running the code from the notebook:
https://github.com/NVIDIA/cuopt-examples/blob/branch-24.11/notebooks/routing/managed_service/cpdptw-reoptmization.ipynbRelevant log output
Environment details
Other/Misc.
Probably the error is found on this function. The task locations are repeated due to the functioning of this method.