Mathematics of Baking
Tim Wichelmann & Jakob Eberhardt
The project is structured as followed:
- models/ includes the developed OPL models
- src/py/HeuristicsProject incudes the implementations of
- Greedy
- local search
- GRASP
- Instance Generator
- src/py/misc includes various scripts for benchmarking and plotting data regarding the Python implementations
- Makefile can be used to automate CPLEX benchmarks
- src/py/HeuristicsProject/data includes example instances of many sizes
- data/report/results.ods includes the benchmarking results used in the report
- In order to run a single instance, we can use
make, which will obtain a optimal solution for the input data specified in theMakefile(default 75 orders). The output is also stored to a.solfile. - To run and benchmark the model for all instances available in src/py/HeuristicsProject/data, we can use
make full. The results will also go to src/py/HeuristicsProject/solutions_cplex.
All runs are configured in the config.dat file located under HeuristicsProject/config/.
In pycharm, we can select one of the run configurations in the top right bar. These include:
- Run Solver which will run the configured instance.
- Generate Data to manually generate an instance (e.g. for CPLEX).
- Generate Plot to create plots based on the solutions.
Alternatively, we can use the following commands to run each script form the command line:
### Run the solver with the given configuration
PYTHONPATH=<path/to/project>/projects/ammm/src/py python3.8 Main.py### Manually generate data
PYTHONPATH=<path/to/project>/ammm/src/py/HeuristicsProject/DataGenerator/Main.py python3.8 Main.py### Create schedule plots
PYTHONPATH=<path/to/project>/ammm/src/py/misc/main.py python3.8 Main.pyAfter configuring the run in the config.dat, we can use the scripts located under misc/ to automatically change input parameters such as maxExecTime or inputDataFile. The output is saved to the respective .sol file.