Closed ammen99 closed 1 year ago
Now we can use the following:
./release/src/crp -l 4 -c 4 -i data/karlsruhe -w travel_time --dump-partition > data/karlsruhe/p4x4inert
This runs only the partitioner and dumps the bitmask on stdout. We save it on a file so that we do not have to recompute the partition on every run - this can be useful when optimizing the customizer or the queries later.
Then we can:
./release/src/crp -l 4 -c 4 -i data/karlsruhe -w travel_time -p data/karlsruhe/p4x4inert --dump-customization > data/karlsruhe/p4x4inert_cust
To run the customizer only and dump the customization matrices on stdout for later use. Finally, to run the algorithm with both phases precomputed:
./release/src/crp -l 4 -c 4 -i data/karlsruhe -w travel_time -p data/karlsruhe/p4x4inert -C data/karlsruhe/p4x4inert_cust -q data/karlsruhe/test
Note that if you want to try the full algorithm, you can still do that, just specify the actual algorithms instead of files:
./release/src/crp -l 4 -c 4 -i data/karlsruhe -w travel_time -p inertial -C dijkstra -q data/karlsruhe/test
Now we can use the following:
This runs only the partitioner and dumps the bitmask on stdout. We save it on a file so that we do not have to recompute the partition on every run - this can be useful when optimizing the customizer or the queries later.
Then we can:
To run the customizer only and dump the customization matrices on stdout for later use. Finally, to run the algorithm with both phases precomputed:
Note that if you want to try the full algorithm, you can still do that, just specify the actual algorithms instead of files: