Cyclopts

The Cyclus Optimization Library

Build & Install

The full process of building and installing Cyclopts can be broken into two sections:

1. building and installing the dependencies
2. building and installing Cyclopts

Installing Dependencies

To install Cyclopts, one must first install:

• boost: http://www.boost.org/
• coin-Cbc(min. v2.5): https://projects.coin-or.org/Cbc

Now, there's an easy way to install these dependencies and a hard way. I'll detail the easy way and gloss over the hard way, which is for more experienced users anyway.

The Easy Way: Using a Package Manager

If you have access to a package manager (e.g. Synaptic for Ubuntu), then the installation process of the dependencies is easy -- you can get them all through the manager. If you've heard of something called apt-getting, then you're golden. If not, might want to get familiar with it if any errors come up that you don't understand.

To perform the following process, we assume that you have root (i.e., super user) access in order to issue "sudo" commands. This process was tested on a fresh install of Ubuntu 12.10. All commands were required to fully install the dependencies.

To install the dependencies the easy way, you'll need to issue the following command:

sudo apt-get install <package>

where you will replace the term <package> with each of the following:

1. g++
2. libboost-dev
3. cmake
4. libbz2-dev
5. coinor-libcoinutils-dev
6. coinor-libosi-dev
7. coinor-libclp-dev
8. coinor-libcbc-dev
9. coinor-libcgl-dev

The Hard Way: Building from Source

The installation instructions to install boost from source are provided on their website.

The direct dependency for Cyclopts is coin-Cbc. It is possible to build this dependency from source, but you may find that you need to install additional dependencies based on the version you're using. See the FindCOIN.cmake file in /src/CMake for some comments.

If you choose to install coin-Cbc in a non-standard location (i.e., some location not in a standard environment variable that cmake checks like PATH and LD_LIBRARY_PATH), you must inform CMake as to its location using the command line argument COIN_ROOT_DIR.

Installing Cyclopts

Assuming you have the dependencies installed correctly, it's pretty straightforward to install Cyclopts. We make the following assumptions in this guide:

1. you have acquired the Cyclopts source code from the repo
2. there is some master directory in which you're placing all Cyclus-related files called .../cyclus
3. you have placed the Cyclopts repository in .../cyclus/cyclopts
4. you have a directory named .../cyclus/install in which you plan to install all Cyclus-related files
5. you have a directory named .../cyclus/cyclopts/build in which you plan to encapsulate all build-related files (they get in the way otherwise)

Under these assumptions and if you used a package manager to install coin-Cbc (i.e. it's installed in a standard location), the Cyclopts building and installation process will look like:

.../cyclus/cyclopts$ python setup.py --prefix=../install

If you have installed coin-Cbc from source or otherwise have it installed in a non-standard location, you should add an appropriate flag

.../cyclus/cyclopts$ python setup.py --prefix=../install --coinRoot=path/to/coin/libraries

You can always investigate the full set of installtion options via:

.../cyclus/cyclopts$ python setup.py -h

Run an Example

An example application that uses the Cyclopts library is provided in the examples folder.

Assuming you have built and installed Cyclopts in the manner described above, you can issue the following commands to build and run the example (remember, we installed Cyclopts in .../cyclus/install).

.../cyclus/cyclopts$ cd examples
.../cyclus/cyclopts/examples$ python setup.py --cycloptsRoot=../../install

Again, if you installed coin in a nonstandard location, you'll also have to inform the setup script.

.../cyclus/cyclopts$ cd examples
.../cyclus/cyclopts/examples$ python setup.py --cycloptsRoot=../../install --coinRoot=path/to/coin/libraries

Some Notes

COIN-OR is the current tool used to solve LPs and MILPs, however, others exist:

1. glpk
2. pyomo
3. symphony