API for accessing E-PICSA climate services

Build with FastAPI

The api requires Python and R runtimes installed. It has been tested with the versions listed below

• Python (3.11)
  https://www.python.org/downloads

• R and Rtools (4.2.3)

  Windows
  https://cran.r-project.org/bin/windows/base/old/4.2.3/
  https://cran.r-project.org/bin/windows/Rtools/rtools42/rtools.html

  Linux
  https://cran.r-project.org/bin/linux/

  Mac (untested)
  https://cran.r-project.org/bin/macosx/

Relevant documentation should also be followed to ensure runtimes can be executed from PATH environment variable.

Environment

Create an environment file from the sample. These will be populated using Pydantic

cp .env.sample .env

Python

The scripts below will create a python virtual environment, activate, install required python and R dependencies and start local server

=== "Windows (powershell)"

python -m venv .venv
.\.venv\Scripts\Activate.ps1
pip install --upgrade -r requirements.txt
uvicorn app.main:app --reload

=== "Linux (bash)"

python -m venv .venv
source .venv/bin/activate
pip install --upgrade -r requirements.txt
uvicorn app.main:app --reload

R

Once installed you will need to call R from an elevated shell to install dependencies

Windows

Rscript install_packages.R
Rscript install_packages_picsa.R

Linux

sudo Rscript install_packages.R
sudo Rscript install_packages_picsa.R

Authorization File

In order to run the package, you will need to add the following service account file to the main repository folder (i.e. the folder where this README.md file is stored):

service-account.json

Once installed, subsequent server starts can skip installation steps

=== "Windows (powershell)"

.\.venv\Scripts\Activate.ps1
uvicorn app.main:app --reload

=== "Linux (bash)"

source .venv/bin/activate
uvicorn app.main:app --reload

The server will start at http://127.0.0.1:8000

pytest

docker compose up --build

This repo contains example workflow to build as a docker image and deploy to google cloud run. See action yaml for details

Pip won't install dependencies Depending on local versions of R and python (as well as operating system) there may be issues when installing certain packages. Recommend attempting install using the requirements_dev.txt file which pins exact versions of packages shown to be compatible with each other, i.e.

pip install --upgrade -r requirements_dev.txt

R Dependency incompatibility Ensure R installed a per prerequisites, verify version via rscript --version.

If facing issues with a specific package it may help to download RStudio, and using the packages tab to check what version of packages are installed and update any indicated within error logs.

Called endpoint method does not exist The library calls methods from various other git repos where code is hosted both in python and R. These are installed during initial setup, but will need reinstallation whenever new versions of the external repos exist.

Simply repeat the steps above to install dependencies from install_packages_picsa.R

Any other issues should be raised on GitHub

This project is licensed under the terms of the MIT license.