CLI to run my graduation degree tests. I've implemented two models for the minimal common substring partition (MCSP) problem described by Blum e Raidl (2016) using Google OR-Tools. The instances are provided by Ferdous e Rahman (2013).
If you want to use commercial solvers like CPLEX, you have to compile it from the source, as explained here. Otherwise, you can install it from pypi package.
pip install --upgrade ortools==9.6.2534I was unable to install ortools in my system after building it. A workaround is to use the command
source for the virtualenv in the building folder.
source /path/to/or-tools-9.6/build/python/venv/bin/activateDependencies can be installed with pip as follows.
pip install -r requirements.txtAs this program uses google spreadsheets API, you need to provide your own authentication credentials.
To generate the client secret file from Google OAuth you can follow this tutorial.
Add your google account to test users in your google cloud project, so you will be able to give this program access to your spreadsheets. To do so, go to your project console > APIs & Services > OAuth consent screen > Test users > Add users.
Before using it, rename or copy the dotenv file to .env and assign the variables the proper values.
| Flags | Description |
|---|---|
| -h, --help | Show this help message and exit |
| -v, --version | Show program's version number and exit |
| --verbose | Show debug messages |
| --local | Don't send send the results to spreadsheets, just print it to stdout |
| --dry-run | Don't run the tests, only log |
| --limit LIMIT | Time limit in milliseconds |
| -m, --models {cb,cs} [{cb,cs} ...] | Defines which model should be used for resolve MCSP problem instances |
| -g, --group {1,2,3,real} | Defines which group of test cases should be used |
| -c, --cases [CASES ...] | Defines which instances should run for testing |
| -n, --num-executions NUM_EXECUTIONS | Define how many times a instance should execute |
| -s, --solvers [{CPLEX,GUROBI,SCIP,CBC} ...] | Defines which solvers should be used |
| --heuristic {chrobak} | Set a heuristic to be used as warm start |
In this case we are running model Common Substring (-m cs) using instance 113 (-c 113) from group 1 (-g 1), applying the chrobak heuristic(--heuristic chrobak) as warmup, using CPLEX as solver (-s CPLEX), and each instance is only running once (-n 1).
python tests_cli.py -m cs -g 1 -c 113 -n 1 -s CPLEX --heuristic chrobak