gfold 0.1.0

G-FOLD: Fuel Optimal Large Divert Guidance Algorithm

G-FOLD Python Generator

This is the Python implementation of the G-FOLD algorithm with code generation capabilities. The problem is described using CVXPY/Python and C/C++ code is generated using CVXPYGen.

Installation

Option 1: Install from source

Prerequisites

You need to install Rust and Eigen for the code generation feature.

Clone the repository:

git clone https://github.com/samutoljamo/g-fold.git
cd g-fold/generator

Install the package in development mode:

pip install -e .

On WSL, make sure you've installed Tkinter (version depends on the python version you're using):

sudo apt-get install python3.12-tk

Option 2: Install from PyPI (coming soon)

pip install gfold

Usage

As a command-line tool

After installation, you can run G-FOLD from the command line:

# Solve the example problem with 100 steps and display graphs
gfold -n 100

# Generate C++ code
gfold -g -n 100 -o output_directory

# Save the plot to a file without displaying it
gfold -n 100 --save-plot --no-plot

As a Python library

from gfold import GFoldSolver
from gfold.visualization import plot_results

# Create a solver with 100 steps
solver = GFoldSolver()

# Solve the problem
solution = solver.solve(verbose=True)
print(f"Final mass: {solution['final_mass']:.2f} kg")

# Plot the results
plot_results(solution, save_path="gfold_plot.png")

# Generate C++ code
solver.generate_code(code_dir="generated_code")

Example scripts

Check the examples directory for more usage examples:

• simple_example.py: Basic usage of the solver and visualization

Development

To set up the development environment:

cd g-fold/generator
pip install -e .