Installation
This document covers a few ways to install esgpull, a necessary first step into being able search and download ESGF datasets.
Python version¶
esgpull only supports python 3.10 and newer.
Supporting lower python versions could be done with future releases, do not hesitate to ask for it.
Setup¶
Installing esgpull is the first step to using it, but not the only one.
Once you have installed the esgpull package, make sure to read this section.
Only a few esgpull commands work out of the box, namely search and convert.
To use the full set of functionalities, you will need to setup a local install with:
Why do I need to install twice ?¶
The reason is that esgpull is prevented from writing anything on disk until installed.
Installing esgpull equates choosing a directory in which it is allowed to write anything it needs to run properly. It also creates all the required files/directories in that directory and fetches some metadata from ESGF that is required to run properly.
Multiple installs¶
It is possible to have multiple installs, which allows using multiple configurations on the same machine.
Installing in a directory which is an existing esgpull install is possible and intended. It allows sharing a single configuration across multiple users.
Deleting an install¶
To delete an esgpull install, it is required that this install is active.
Deleting does not rm any file, it only removes the option to use this install.
Installing with conda / mamba¶
To install esgpull in a new conda environment, run:
For mamba users:
You can start using esgpull after activating the environment:
Install with pip¶
Make sure your python version meets the requirements (>=3.10), then you can run:
Install from source¶
Esgpull is developed and maintained on GitHub, you can clone the public repository with:
And then install with pip: