-
Notifications
You must be signed in to change notification settings - Fork 26
Run HyperMapper
Luigi Nardi edited this page Dec 20, 2019
·
1 revision
Before running HyperMapper first setup a json configuration file. Follow this example: example_scenarios/spatial/BlackScholes_senario.json
cd hypermapper;
python3 scripts/pareto_based_active_learning.py example_scenarios/spatial/BlackScholes_senario.json
This script computes a Pareto from the csv data files specified in the json output_data_file field and returns the Pareto in the file specified in the json output_pareto_file field.
python3 scripts/compute_pareto.py example_scenarios/spatial/BlackScholes_scenario.json
Need the following files:
- Approximated Pareto: this is the result of your design-space exploration under the form of a csv file.
- Output result of the design-space exploration this is the file with all the samples explored during the search, under the form of a csv file. In practice two files are generated:
- A first file where the view is restricted to [0, 1] on the axis where the percentage option was enabled in the json.
- A second file with the prefix "all_" where all the samples are plot without restriction.
- The search space file: this is the file that contains the definition of the space to be searched. Example: example_scenarios/spatial/BlackScholes_search_space.pcs.
- [Optional] Pdf output file: this is the image created by this script. If not provided the output.pdf will be generated.
- [Optional] The real Pareto: this is the actual Pareto (not always available) under the form of a csv file.
python3 scripts/plot_dse.py example_scenarios/spatial/BlackScholes_senario.json
You can run one script that does the 3 steps in one: active learning + compute Pareto + plot.
Follow this example: example_scenarios/spatial/BlackScholes_senario.json
cd hypermapper;
python3 scripts/hypermapper.py example_scenarios/spatial/BlackScholes_senario.json