development_environment_setup.md

Development Environment Setup

Dependencies

Operating system requirements: Mac, Linux, and Windows (WSL 2 only)

1. Clone the repository
2. Install dependencies:

./script/bootstrap

3. Restart your command prompt.

IDE

We use Visual Studio Code (VS Code) for development.

Model

Navigate to the model folder.

Unpack the data into the data/ folder. This data was not included in the repository because it is proprietary data belonging to CSI Pacific.

The data/ file structure should look like this:

data/
- <pole_labels_file_name>.csv
- <boot_labels_file_name>.csv
- data/

Add the pole and boot labels file paths to /src/data-processing/config.py

Activate the conda environment:

conda activate model-env 

Test that Jupyter Notebook works, run the following command and open a notebook in notebooks/.

$ jupyter notebook

Open the folder model in VS Code. (Only open VS Code in this folder when working on the model!)

Edit the model/.vscode/settings.json file as follows:

{
    "python.autoComplete.extraPaths": ["<path to repo root>/model/src"],
    "python.pythonPath": "<path to user home directory>/miniconda3/envs/model-env/bin/python",
    "jupyter.jupyterServerType": "local",
}

As an alternative to using the jupyter notebook command, you should be able to run and use Jupyter notebooks in VS Code.

• Refer here for troubleshooting.

• Note however that plotting with Matplotlib may be buggy.