index.md

Anaconda and conda

One attribute of Python that makes it a great language for science is its abundance of packages (numpy! scipy! pandas! xarray! pymt!). Package management can be difficult, though, especially when a typical Python installation contains dozens of packages.

This is where the conda package manager is handy, and a primary reason why CSDMS uses (and we recommend) the Anaconda Python distribution (now called Anaconda Individual Edition).

With conda, you can:

• list
• install
• update
• remove

packages from a Python installation. conda ensures that the packages work together without conflict.

Because much of our work in ESPIn takes place on the CSDMS JupyterHub, we don't spend much time on conda here; however, we recommend installing Anaconda on your computer so that you can use all of the ESPIn course material locally.

Further, to ensure you have all the correct packages needed to use the course material, we ask that you set up an environment, an independent Python installation managed by conda. To do so, we'll need an environment file from the ESPIn repository.

In the "Getting things from elsewhere" section of the shell lesson, we downloaded the ESPIn repository as a zip archive and uncompressed it in our Desktop directory. In a terminal, change to this directory and view the file environment.yaml with cat:

$ cd ~/Desktop/espin-main
$ cat environment.yaml
# A conda environment for ESPIn lessons.
#
# Usage:
#   $ conda env create --file=environment.yaml
#   $ source activate espin

name: espin
channels:
  - conda-forge
dependencies:
  - python =3
  - numpy
  - scipy
  - pandas
  - notebook
  - bmipy
  - pymt >=1.1
  - landlab >=2.0

The environment file lists all the packages needed to run the course material. If a package has a dependency not listed (e.g., pymt is built on xarray), conda finds a compatible package version for you.

To create the environment, type:

$ conda env create --file=environment.yaml

Once the environment has been created, type

$ source activate espin

to make this environment current.

Later, when finished using the environment, type

$ conda deactivate

to return to the base environment, and

$ conda remove -n espin --all

to fully remove the environment from your Anaconda installation.

Summary

While Python is installed with most operating systems, or can be downloaded and installed from source, CSDMS recommends the use of Anaconda because of its conda package management system.

This table summarizes conda concepts covered in this section:

Concept	Description
package manager	a tool for managing Python packages, ensuring compatibility
environment	an independent, isolated Python installation managed by conda
environment file	a file specifying the packages that make a conda environment

This table summarizes the conda subcommands used in this section:

Subcommand	Description
list	lists all packages installed in an environment
create	make a new environment from a file
install	adds a new package into an environment
remove	uninstalls a package from an environment
update	gets newer versions of packages in an environment

This table summarizes shell commands used in this section:

Command	Description
source	runs a script

Resources

• The conda User guide
• A conda command cheatsheet from Anaconda