freva 2206.0.10

Free Evaluation and Analysis Framework (Freva)

Free Evaluation System Framework

What is Freva ?

Freva, the free evaluation system framework, is a data search and analysis platform developed by the atmospheric science community for the atmospheric science community. With help of Freva researchers can:

• quickly and intuitively search for data stored at typical data centers that host many datasets.
• create a common interface for user defined data analysis tools.
• apply data analysis tools in a reproducible manner.

Data analysis is realised by user developed data analysis plugins. These plugins are code agnostic, meaning that users don't have to rewrite the core of their plugins to make them work with Freva. All that Freva does is providing a user interface for the plugins.

Currently Freva comes in three different flavours:

• a python module that allows the usage of Freva in python environments, like jupyter notebooks
• a command line interface (cli) that allows using Freva from the command lines and shell scripts.
• a web user interface (web-ui)

Where can I find the Freva user documentation?

A more detailed overview on the usage of freva can be found on the freva user documentation page

How can I install Freva at my institution?

Deployment is realised via a dedicated repository that holds code to set up the command line and web user interface as well as all services. To deploy the system in production mode consult deployment docs.

How can I set up a local version for development?

To start development with freva clone the repository and its submodules:

git clone --recursive https://gitlab.dkrz.de/freva/evaluation_system.git

A basic local development setup can be created using Docker and docker-compose (Linux users need to install it separately).

This also requires that the .envrc file is sourced.

docker-compose up -d

Dummy data can be injected into a running docker-compose environment with make dummy-data. This will add some example files into solr and run an example plugin a few times to add some history data.

When finished, tear down the environment with

docker-compose down

Creating a dedicated anaconda dev environment

We recommend using anaconda to install all packages that are needed for development. Here we assume that you have a working anaconda version per-installed on your local computer. To install the dev environment simply use the following command:

conda env create -f dev-environment.yml
source .envrc

This will automatically set environment variables needed for development. The freshly installed environment can be activated:

conda activate freva-dev

The conda environment can be deactivated using the following command:

conda deactivate

Note: The conda install command can be slow. If you want to speed up the installation of the environment we recommend to install the mamba package in the anaconda base environment and use the mamba command to create the environment:

conda install mamba
mamba env create -f dev-environment.yml
source .envrc

Installing the python package

Use the pip install command to install the actual python core packages into your activated environment:

pip install -e .[test]

The -e flag will link the source code into your python environment, which can be useful for development purpose.

Running tests and creating a test coverage report

The system can be tested with a Makefile. To run the tests and generate a simple test coverage report simply use the make command:

make test

The linter testing can be applied by:

make lint