Ipopt is a state-of-the-art optimization solver for nonlinear optimization problems. Unfortunately, the only interface available is C/C++ or Fortran. As Python becomes more popular these days, a connector to python is necessary. This project consists several major parts:

1. A simple connector using python list to pass the value for users don't want to get things done quickly.

2. A complicated version of pyipopt that using numpy data types from users who wants to do high performance computation.

3. With the help of nlpy, now you can directly manipulate AMPL nl file in python directly and send it to ipopt with one or two lines of code.

4. A connector to openopt (DONE by openopt community now)

5. Pyipopt support multiple instances of problems. Aka, you can use one module and solve multiple problem instances.

Pyipopt is reasonably robust according to my own test.

This work is sponsored by my advisor Dr. Yixin Chen's Early Career Young Investigator Award from the Department of Energy and is done at Washington University.