MATSP

MATSP is an ATSP instance solver based on a Memetic Algorithm (MA) approach.

Environment Creation

Before executing the application, you must create a Python environment and install the requirements.

# (First Option) Using the simplified commands provided with Make.
make create-venv
make install-requirements

# (Second Option) Using the standard Python commands.
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Usage Examples

Running a single configuration file:

# (First Option) Using the simplified commands provided with Make.
make run CONFIG="resources/configuration.example.yml"

# (Second Option) Using the standard Python commands.
source venv/bin/activate && python3 -u src/main.py --config "resources/configuration.example.yml"

Running all configuration files in a directory:

# (First Option) Using the simplified commands provided with Make.
make run CONFIG="resources/configurations/"

# (Second Option) Using the standard Python commands.
source venv/bin/activate && python3 -u src/main.py --config "resources/configurations/"

All YAML configuration files are parsed and executed automatically. A single YAML file may contain multiple configurations.

The output includes:

• a Markdown summary for each configuration;
• an individual convergence plot for each configuration; and
• a combined comparison plot containing:
  • convergence curves (with and without standard deviation);
  • best costs; and
  • average execution times.

The combined comparison plot is generated only when more than one configuration is executed.

Configuration Naming

When executing multiple configuration files from a directory, file ordering matters.

Avoid naming files like this:

config_1, config_2, ..., config_10, config_11

They will be ordered lexicographically:

config_1, config_10, config_11, config_2, ...

This happens because filenames are compared character by character.

Instead, use zero-padded numbering, which is the standard convention in scientific computing, datasets, simulation pipelines, rendering systems, logging infrastructures, and backup systems.

config_01, config_02, config_03, ..., config_10, config_11, config_12

Transparency Commitment

Artificial Intelligence (AI) tools were used during the development of this project for the following purposes:

During development

• evaluating alternative implementations and designs;
• debugging problems with unknown solutions;
• designing the matrix extraction algorithm;
• learning standard solutions and Python practices, such as protocols (PEP 544); and
• designing the YAML configuration expansion system.

After development

• optimizing some algorithms and implementations; and
• reviewing the project for inconsistencies and typographical errors.

Disclaimer

MATSP is an educational project.

It is not intended for production environments, nor is it expected to provide state-of-the-art performance for large-scale industrial ATSP instances.

Credits

• González Tesoro, Joaquín
• Roumec, Iñaki