Running the BEAST requires:
In turn, Astropy depends on other packages for optional features. From these you will need:
hdf5 to read/write
Tableobjects from/to HDF5 files.
You will also need:
PyTables to manage large amounts of data.
One easy way to obtain the above is through the AstroConda Python stack:
First install Miniconda which contains the conda package manager. Once Miniconda is installed, you can use the
condacommand to install any other packages and create environments, etc.
Setup the AstroConda Channel
$ conda config --add channels http://ssb.stsci.edu/astroconda
Install AstroConda with Python 3 (recommended)
$ conda create -n astroconda stsci
Make sure that the
hdf5packages are installed
$ conda install -n astroconda pytables $ conda install -n astroconda hdf5
Installing the BEAST¶
In addition to installing the code, library files also need to be installed. See BEAST Library Files.
numpy must be installed before installing the beast. This package
is needed as part of the install process, not just as a prerequisite for the
beast can be installed using pip
$ pip install beast
if you already have an older version installed
$ pip install --upgrade beast
from the master trunk on the repository, considered developmental code
$ pip install git+https://github.com/BEAST-Fitting/beast.git
If you are happy with your current environment,
beast can also be installed from
the source code in the normal python fashion after cloning from the git repo or
downloading from Github
$ python setup.py install
If you are using conda, you may wish to create a conda environment with the dependencies before doing the install
$ conda env create -n beast --file conda-environment.yml $ conda activate beast $ python setup.py install
If you would like to modify beast, you may want to use links instead of installing, which is best done by replacing the last line with
$ python setup.py develop
Installation for Production Runs¶
Using a dedicated conda environment for production BEAST runs may be desirable. Such an environment provides a way to ensure that production runs are reproducible by fixing the versions of all the software used. The instructions below assume that the astroconda channel is being used.
Create a conda environment. Here we name it to include the BEAST version.
$ conda create -n beast_v1.3 python=3.6
Activate the environment after all the packages are finished installing.
$ source activate beast_v1.3
Install dependencies using conda (better for speed)
$ conda install astropy scipy h5py matplotlib cython
Next, install the BEAST. You have three options:
Option 1: Use pip to install the production version of the beast (currently v1.3)
$ pip install beast==1.3
Option 2: Get the latest production branch, which can be ahead of pipy version
$ pip install git+https://github.com/BEAST-Fittingemail@example.com
Option 3: If you’ll be doing development, fork the beast (as described
navigate into the first
beast folder, and do this command. Any changes
you make will be immediately reflected in your calls to the BEAST code. Note that
you can make separate environments for development and production modes.
$ python setup.py develop
The BEAST production version is now ready for use. Note, you need to activate this conda environment every time you want to use this installed version.
BEAST Library Files¶
For the BEAST to work properly, you need to place a set of files in a directory. These files contain information related to filters, stellar atmospheres, and in the future stellar evolution models.
The required library files can be manually acquired from:
Note that the archive at this link contains a folder called
contents of this folder are the library files required by beast. It is these
files that need to be placed within (any of) the possible locations specified in Location.
Alternatively, after installing the BEAST, run the following script and the library files will be downloaded into the location specified in Location
$ beast get_libfiles
There are 2 possible locations for the required library files. The possible locations are (in the order the code will search for them):
In a directory designated by a BEAST_LIBS environment variable
.beastdirectory in the home directory of the current user (ie,
~/.beast); this is usually the easiest and will be automatically created if it doesn’t exist.