Location for code related to calculating the unimpaired flow and applying different water allocation approaches to determine how the flow would change based on each approach. This code was used to support the analysis for a paper entitled "Evaluating approaches for protecting environmental flows in decentralized water management systems" that is currently in preparation for publication in a peer-reviewed journal.
The code is written as jupyter notebooks and is based on a series of python files written by Jesse Rowles ([email protected]). Those python files are included as reference files in this repo in the XX folder. Note that these instructions and the environment are designed to run the jupyter notebook files and not the python files, so additional python packages may be needed to run that code. See the readme in that folder for more details.
For use on windows machines, use the miniconda package manager.
Download Miniconda from https://conda.io/miniconda.html. Make sure to install the Python3 and 64-bit version. Installation instructions for windows is here https://docs.anaconda.com/miniconda/miniconda-install/
After it is installed, click on the start menu on the lower left of the screen to see all programs installed. You should see a "Anaconda3 (64 bit)" folder and inside that folder should be a program called "Anaconda Prompt". Start that program. A new code window will open up. You can check the installation by typing conda --version. It is also adviceable to run conda update conda before you start.
Test whether Git is installed on your system by typing in the same Anaconda Prompt window:
git --version
If no version number is returned, install Git from https://git-scm.com/.
We need a directory where the project will live. On a windows machine,you may use a directory like "C:\Users\USERNAME\Documents\workspace" where USERNAME is your user name for your computer. Some of the files are quite large so make sure you have a several GB of free space. I will refer to this location as PATH. In the "Anaconda Prompt" window, navigate to this location by using typing in the correct drive (most people it will be "C:" but for me it is the "E:" drive) and then use the "cd" notation for "change directory". Type in "cd" a space, and your PATH name. For me it is
E:
cd E:\Users\kklausmeyer\Documents\workspace
For you it may be different. Now the prompt should read "(base) PATH>" where PATH is your path.
In order to create a new directory instream-flow-management
mkdir instream-flow-management
mkdir stands for make directory. Now navigate to that directory by using the cd notation.
cd instream-flow-management
Now the prompt should read "(base) PATH\instream-flow-management>" where PATH is your path.
From within the instream-flow-management directory run:
git clone https://github.com/tnc-ca-geo/instream-flow-management.git
This will create a new folder or directory with the same name "instream-flow-management" inside the "instream-flow-management" folder. Moving forward, we will refer to the two folders as the outer flow-update directory and the inner flow-update directory.
If an (inner) instream-flow-management directory already exists within the (outer) instream-flow-management directory, you will get an error message. In this case change using into the (inner) instream-flow-management directory and run
cd instream-flow-management
git pull origin
Repeat this whenever you start working with the project to get the latest version.
It should (hopefully) be very easy to create the environment from the conda_environment.yml file in the repository. Change into the (inner) instream-flow-management directory, make sure conda_environments.yml is there (use the dir command) and run.
conda env create --file conda_environment.yml
This will take awhile and download a bunch of dependencies, eventually it should exit wthout an error message.
If you want to learn how to create an environment manually and how to add depencies to that environment see https://conda.io/docs/.
Please keep the conda_environment.yml file up to date if you introduce new dependencies into the project. As the project evolves other programmers might introduce new dependencies as well. You should update the environment after pulling a newer version of the project from Github
conda env update --file conda_environment.yml
Type
conda activate instream-flow-management
into a cmd shell in order to activate the environment.
Note: You need to do this whenever you open a new cmd shell. Environment activations are bound to the shell. Different shells may have different environments activated. If you use any integrated development environment such as IDLE, there are project-level settings that determine which environment will be used if you press the play button.
Some of the packages are not available to install via conda forge and require the use of pip to install. Once the environment is active, type the following commands:
pip install -U kaleido
pip install loess
This project uses Jupyter notebooks to review output in real time. When you set up the environment, you need to create a new kernel associated with the environment. This command will do that:
python -m ipykernel install --user --name instream-flow-management --display-name "Python (instream-flow-management)"
Note If you have trouble with a Kernal Error in jupyter notebook, try installing pypiwin32. Do this at the conda prompt and type
pip install pypiwin32
When returning to the project to update the code, folow these steps:
- Launch the "Anaconda Prompt" from your start menu (on a windows machine)
- Use the cd notation to navigate to the inner directory of your working folder
cd PATH\instream-flow-management\instream-flow-management
or
cd C:\Users\kklausmeyer\Documents\workspace\instream-flow-management\instream-flow-management
- (optional) Update the environment:
conda env update --file conda_environment.yml
- Activate the environment
conda activate instream-flow-management
- Get the latest version of the code from GitHub
git pull origin
- Start Jupyter notebook
jupyter notebook
Here is a nice summary of the basic git commands to sync changes to the master repo https://confluence.atlassian.com/bitbucketserver/basic-git-commands-776639767.html