SOSPy 0.1.0

Python version of MATLAB SOSTOOLS, a toolbox for sum-of-squares optimization

This package is a Python version of MATLAB TOOLBOX -- SOSTOOLS.

This package contains functions based on sympy package. However, in case of a large number of decision variables, sympy may take a long time to parse the data.

Demos are in the folder SOSPy_demos.

Updates in version 0.1.11:

• Fix bugs in findsos().
• Warning: for solvers in sossolve(), although SCS is installed and can be used by cvxpy, it may not work properly.

Updates in version 0.1.10:

• Print latex form of the solution when calling sosgetsol().
• Adjust the logic of changing solvers when one encounters failure.

Updates in version 0.1.9:

• Fixed some bugs
• When creating an SOS program, the code searches for and prints out all installed and available solvers.
• We have added cvxpy as a solver. As a result, users can now utilize other solvers through cvxpy.
• The solvers we currently support are mosek and cvxpy.
• It's worth noting that since cvxpy acts as a parser, certain results cannot be read after executing with cvxpy. This includes details like the iteration number, primal infeasible, and dual infeasible.

Contributors:

• James Anderson, email: [e-mail]
• Leonardo Felipe Toso, email: [e-mail]
• Zhe Mo, email: [e-mail]