Predictive Services Daily WIMS Feed

Data feed for National Predictive Services program.

Cron lambda deployed to AWS US-West-2 via Serverless used to process and upload daily WIMS output to the web.

Setup

Before getting started in the code, we'll need to set up a few dependencies.

Git and Github

Github is where the code is hosted and git is used for version control. Most computers have git installed but you'll need to set up a GitHub account to push any changes you make to the repository. A great set of instructions for setting up this step is found at the GitHub Quickstart.

Node and Serverless

These dependencies are used for deploying our code and creating our AWS resources. You don't need to do this step if you have no intention of ever deploying the code "live".

Install Node 16.15.0 LTS via the node website. Once Node is installed, run the following to install Serverless:

npm install -g serverless

NOTE: THERE IS SOME CONFIGURATION REQUIRED WITH SERVERLESS AND AWS THAT IS BEYOND THIS README. WHEN/IF THE CODE IS HANDED-OFF, I AM HAPPY TO ASSIST ANY AGENCY STAFF IN GETTING THIS SETUP, ANYTIME

Pyenv and Poetry

Pyenv is used to manage our python versions while poetry is used to manage our python dependencies.

To install pyenv, see the PyENV installation guide. Make sure to follow the instructions closely! After you've installed pyenv, use it to install the python version required:

pyenv install 3.8.6

Now go ahead and install poetry (which is a lot easier):

curl -sSL https://raw.githubusercontent.com/python-poetry/poetry/master/get-poetry.py | python -

With pyenv and poetry installed on your computer, navigate into this repo and run:

pyenv local 3.8.6
poetry install

This will make sure we use the pyenv we just created and then install all needed dependencies in a local environment.

Finally, we can ensure we activate our local environment and run our tests:

poetry shell
pytest

Making code changes

work in progress

Deploying

Assuming we've configured everything correctly for AWS/Serverless, it's as simple as:

yarn deploy