Data and code for Hawaii version of SWITCH
##INSTALLATION
###INSTALL PYTHON AND PYOMO
Python: On recent Mac or Linux systems, a suitable version of Python should already be installed (SWITCH needs one of the 2.7 versions). On Windows, you should download the binary installer from https://www.python.org/downloads/windows/ . If installing on Windows, choose the options to install pip and "Add Python.exe to path".
Pyomo: Once Python is installed go to a terminal window (Terminal.app on a Mac; Windows-R, then cmd on Windows). Then on Mac or Linux execute "sudo -H pip install pyomo". On Windows execute "pip install pyomo".
###INSTALL A SOLVER
SWITCH uses Pyomo to create standard matrices defining the numerical optimization model to be solved. Then it uses standard solvers to solve these models.
CPLEX or GUROBI are high-performance solvers which are available from their developers at no cost for academic users. GLPK is an open-source solver which is free for any user.
On Linux, glpk can be installed via
sudo yum install -y glpk glpk-utils
or
sudo apt-get install -y glpk glpk-utils
On a Mac, the easiest way to install glpk and other unix-style software is via the Homebrew package manager. If you don't want to install Homebrew, you can replace the following steps with instructions from http://hichenwang.blogspot.com/2011/08/fw-installing-glpk-on-mac.html .
Install homebrew package manager by typing the following command in a Terminal window (more details at brew.sh):
ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"
Press Return when prompted, and enter your password when prompted.
Next, you can install glpk itself by typing the following commands in a Terminal window:
brew install git
brew install homebrew/science/glpk
On Windows, glpk can be installed as follows:
- Download the Windows version of glpk from http://sourceforge.net/projects/winglpk/ .
- Open the .zip file you downloaded and look in the W32 or W64 folder (depending whether you have a 32-bit or 64-bit version of Windows).
- Copy glpsol.exe and glpk_n_nn.dll from this folder to C:\Python27\Scripts . (n_nn is the glpk version number, e.g., 4_57)
###INSTALL SWITCH
It is recommended that you install the git command line tool on your system and then follow Option 1 below. Alternatively you can install the latest version by following Option 2 below.
####Option 1
In a terminal window, use the cd command to switch to the folder where you want to install SWITCH-Hawaii. Then execute "git clone https://github.com/switch-model/switch-hawaii-studies.git"
If you want to use a previous version of the model and data, you should checkout the version you want from the repository you have just created. Do this with a command like this:
git checkout <version>
The current options for <version> are v2016-01-15-data and v2016-01-28. You can skip this command or use
git checkout master to use the latest version of SWITCH.
Please note: versions of SWITCH-Hawaii from before 2016-02-03 are currently only compatible with Pyomo 4.1, the cplex solver and a Mac or Linux system (not Windows). Please contact Matthias Fripp at UH ([email protected]) if you would like to run earlier versions of SWITCH-Hawaii in a different environment than this.
On a Mac or Linux system, execute these commands:
cd switch-hawaii-studies/models/rps
./install_switch.sh
On Windows, execute these commands:
cd switch-hawaii-studies\models\rps
[then copy the "git clone ..." commands from install_switch.sh and run them from the command line]
####Option 2
Download the repository from https://github.com/switch-model/switch-hawaii-studies/archive/master.zip. Copy the "switch-hawaii-studies-master" folder from this zip archive to a suitable location (e.g., My Documents) and then rename it to switch-hawaii-studies. Make a note of the name and location of the folder you have created.
Download the repository from https://github.com/switch-model/switch/archive/master.zip. Copy the "switch-master" folder from inside this zip archive into the "models/rps" folder within the "switch-hawaii-studies" folder that you just created. Rename "switch-master" to "switch".
Download the repository from https://github.com/switch-model/switch-hawaii-core/archive/master.zip. Copy the "switch-hawaii-core-master" folder from inside this zip archive into the "models/rps" folder within the "switch-hawaii-studies" folder. Rename "switch-hawaii-core-master" to "switch-hawaii-core".
Whether you followed Option 1 or Option 2, you should now have a directory structure like this (it will have other files too, but these are most of the important ones):
switch-hawaii-studies/
data/
models/
rps/
inputs/
inputs_tiny/
scenarios_to_run.txt
solve.py
switch/
switch_mod/
switch-hawaii-core/
###RUN SWITCH
In a terminal window, use the cd command to get to the switch-hawaii-studies/models/rps directory you created earlier. Then execute this command to run the model:
python solve.py
Inputs will be read from the inputs directory, and outputs will be written to the outputs directory.
"scenarios_to_run.txt" defines the scenarios that should be run. "completed_scenarios.txt" is a list of
scenarios that have already been run. To re-run a scenario that has already been run, you can remove it
from "completed_scenarios.txt" and run python solve.py again. Or you can just execute
python solve.py --scenario <scenario name>. You can also run ad hoc scenarios by specifying
python solve.py --scenario_name <new_scenario> followed (optionally) by command line arguments
to change the scenario. You can see examples of command-line arguments in scenarios_to_run.txt
For testing purposes, it is helpful to use the "inputs_tiny" directory, via a command like this:
python solve.py --scenario_name test --inputs inputs_tiny
##SUPPORT If you need help installing or running SWITCH-Hawaii or defining new scenarios, please contact Matthias Fripp at the University of Hawaii at [email protected].