Installation¶
There are two options for installing elm:
Install with Conda¶
Conda is a package manager backed by Continuum Analytics, Inc. Notable features include first-class support for Python and R software that depends on C extensions, cross-platform support for Windows/Mac OSX/Linux, and standalone packages that are easy to distribute and deploy.
Stable¶
The “stable” release is more thoroughly tested, but does not include the latest experimental features:
conda create -c elm -c conda-forge -c ioam -c scitools --name earth-env elm earthio
The above command creates a conda environment with the latest stable releases of elm and earthio installed. To begin using, activate the environment:
source activate earth-env
If you encounter any issues with installation of the latest release from conda or source, then raise a github issue in the elm repo here or email psteinberg [at] continuum [dot] io.
Development¶
The “development” releases are less stable, but include newer features:
conda create -c elm/label/dev -c conda-forge -c ioam -c scitools/label/dev --name earth-env-dev python=3.5 elm earthio
Like for the stable release, activate the environment to begin using elm:
source activate earth-env-dev
Install from Source¶
Installing elm from source is recommended if you want to develop elm features and iterate rapidly over your code changes. To install elm from source:
git clone https://github.com/ContinuumIO/elm
cd elm
export ELM_EXAMPLE_DATA_PATH=~/elm-data
PYTHON_TEST_VERSION=3.5 EARTHIO_INSTALL_METHOD=conda . build_elm_env.sh
Verify the install with:
python -c "from earthio import *;from elm import *"
You may want to add the following line to your .bashrc (or equivalent shell config) to avoid >1 download of the test data and have the test data discovered by py.test:
export ELM_EXAMPLE_DATA_PATH=~/elm-data
Do the tutorials and examples: