vdx 0.0.3

Python package for index tracking in CasADi

vdx

vdx is the python implementation of the matlab vdx package. It is built on top of the modeling and automatic differentiation tool CasADi. It is used to keep track of indexed variables e.g. x_i∈ Rⁿ for i in 1,...N, as subsets of bigger vectors. This is particularly useful for writing direct transcription optimal control problems such as those found in nosnoc_py.

Installing

As this software is still in development to install the package, clone this repo and use

pip install <path to vdx>

Use

The basic interface for vdx is the NLP object which contains 1 scalar objective f and 3 vectors w (the primal decision variables, g the constraints, and p problem parameters.

To start do:

import casadi as ca
from vdx import NLP, Primal, Parameter, Constraint
nlp = NLP(symbolic_type=ca.SX)

nlp.w.x[range(1,11), range(1,3), range(1,3)] = Primal("x", 4, lb=0.0) # Creates x_i_j_k with i=1,...10, j=1,2, k=1,2