This source code accompanies the paper:
- Decision Diagrams for Solving Traveling Salesman Problems with Pickup and Delivery in Real Time
All source is licensed under the ZIB Academic License, and can only be used
or referenced for academic purposes. There is no warranty and your mileage
may vary. See the file LICENSE.
Building tsppd-dd requires go, and has only been tested on Linux. To build
it, run go build. If that succeeds, you should be able to specify an input file and a diagram form.
./tsppd-dd -input <input json file> -form <form>
If -width is not specified, the resulting diagram will be exact. Otherwise that width controls the restriction and relaxation diagram width. Relaxation and inference duals are specified using the -relax and -infer flags, respectively. For instance:
./tsppd-dd \
-input grubhub-09-4.json \
-form sequential \
-width 5 \
-infer ap \
-batch 10 \
-verbosity 1
instance size form infer relax ordering width batch workers clock cpu primal optimal nodes fails
======================================================================================================================================================
grubhub-09-4 20 sequential ap none 5 10 1 0.001 0.001 8169 false 1 0
grubhub-09-4 20 sequential ap none 5 10 1 0.004 0.004 7543 false 2 0
grubhub-09-4 20 sequential ap none 5 10 1 0.013 0.014 7498 false 42 1
grubhub-09-4 20 sequential ap none 5 10 1 0.014 0.015 7429 false 51 2
grubhub-09-4 20 sequential ap none 5 10 1 0.016 0.017 7156 false 76 8
grubhub-09-4 20 sequential ap none 5 10 1 0.039 0.043 7078 false 270 581
grubhub-09-4 20 sequential ap none 5 10 1 0.117 0.132 7078 true 1055 2379