Installation
AtomDB can be installed with pip. In a virtual environment, simply run the
following command to install AtomDB and its dependencies:
pip install .
If you intend to build the Sphinx documentation, you can also install AtomDB
with the optional dependencies required to do so by appending the [doc] tag
to the install target:
pip install .[doc]
Here is a full example installation:
[NOTE: At this stage in development you must use option C]
# Create a virtual environment in `~/example`.
# Feel free to change the path.
python3 -m venv ~/example
# Activate the virtual environment.
source ~/example/bin/activate
# Then run one of the following:
# A) Install the stable release in the venv `example`.
pip3 install atomdb
# B) For developers, install an alpha or beta pre-release
# (only do this if you understand the implications.)
pip install --pre atomdb
# C) Install the latest git revision
# (only do this if you understand the implications.)
pip install -e .
Setup
If AtomDB is installed in a directory for which you do not have write
permissions, or if you’d rather store your AtomDB database in another directory
(or you already have one), then you can set the ATOMDB_DATAPATH environment
variable to the directory of your choice; either a non-existant directory, or an
existing AtomDB database directory.
On release, AtomDB should come with compiled database files pre-installed in the
default ATOMDB_DATAPATH. Raw data files containing the output of electronic
structure computations will be hosted in another Git repo (AtomDB-Data) using
LFS. If you are a developer working on compiling new database entries or writing
new datasets, then you can set ATOMDB_DATAPATH to the location of the
AtomDB-Data repo on your computer.