Canal block optimization

Canal blocking computation in Dosan, Indonesia.

Installation

1. Create a new environment within Conda with minimal packages (Python 2 version not tested recently, but should work similarly):

Ubuntu 18.04 LTS, Python 3

conda create -n [name of environment] -c conda-forge python=3 numpy scipy rasterio fipy matplotlib pandas xlrd

Windows 10, Python 3

conda create -n [name of environment] python=3 fipy rasterio pandas xlrd

If you're interested in installing it alongside Spyder, only the python 3.6 version works as of March, 2020.

conda create -n [name of environment] python=3.6 fipy rasterio

Optional packages:

• spyder: IDE for Python. Can be added to the list above.
• deap: For Genetic Algorithm implementation. Can be added to the list above.
• simanneal: For Simulated Annealing implementation. Cannot be added to the list above and must be installed separately. Instructions here.

2. Don't forget to activate the new environment!

3. Clone or download this repository.

How to use?

Choose to work either in the Python2 or the Python3 folders. The Python3 folder will be updated, but the Python2 snapshot should work.

There are 3 alternative ways to run the code:

• main.py runs the general computation without any optimization algorithm. It still allows for more than one iteration.
• ga_try_2.py runs the Genetic Algorithm version.
• simanneal_try.pyruns the Simulated Annealing version.

The code needs the following data to run:

• A DEM as a raster image.
• A canal network as a raster image.
• A soil type map as a raster image.
• A soil depth map as a raster image.

The data must be stored in the path "data/" relative to the .py files.