This directory contains the code for the nonlinear modeling benchmarks described in Section 3. 

The proof-of-concept implementation in Julia of the nonlinear modeling features is in julia/nlp.jl. 

To run the benchmarks (inside corresponding directory):

Julia: Run "julia nlbenchmark.jl" to execute the benchmark. This requires JuMP to be installed (see JuMPSection2 directory).

Pyomo (version 3.4): Run "pyomo2nl <model> <datafile>", where <model> is clnlbeam.py or cont5_1.py, with corresponding <datafile>.
E.g. "pyomo2nl clnlbeam.py clnlbeam-5000.dat" runs the clnlbeam-50 reported in Table 4. Execution time is recorded by the line: "[   22.97] Pyomo Finished"

YALMIP (release 20130322, MATLAB R2012b): Run the "doyalmip.m" script.

AMPL (version 20081120): Follow the instructions at the top of main.c to compile the driver. Run "julia benchscript.jl". The ampl executable must be available in the path.