An introduction to the Exoplanet Characterization Toolkit (ExoCTK) package and Web application
The following are instructions on how to install the exoctk
package for both users and contributors. The exoctk
repository provides a conda
environment containing all of the dependencies needed to install and execute the exoctk
software.
You must first have a working installation of anaconda
or miniconda
for Python 3. If you do not yet have this on your system, you can visit the following links for download and installation instructions:
Anaconda <https://www.anaconda.com/download/>
_Miniconda <https://conda.io/en/latest/miniconda.html>
_
To obtain the exoctk
package with the necessary environment files, clone the repository directly from GitHub:
::
git clone https://github.com/ExoCTK/exoctk.git
ExoCTK relies on a large collection of data. Some tools may be run entirely without this data, but light curve fitting, observation planning, and our model grid interfaces require this data. The full data package is 12GB, and will be available from MAST, but we could not arrange this in advance of the workshop. If this is too large to fit on your machine, we have a stripped down set of data that can get you through this workshop, available on dropbox .
After downloading the data in some form, you will have a directory structure
with exoctk_data
at the top level, and subdirectories that correspond to
module specific data, like groups_integrations
, fortney
, generic
,
and so on. Export an environment variable to point to this top level.
::
export EXOCTK_DATA='/path/to/your/exoctk_data/'
This can be a command you run each time you use the package, or added to a
.bashrc
or .bash_profile
.
You can install the ExoCTK conda
environment via the env/environment-<PYTHON_VERSION>.yml
files (relative to the parent directory of where the repository was installed). Note that there are separate environment files for each version of python
that exoctk
supports. First, one should ensure that their version of conda
is up to date:
::
conda update conda
Next, one should activate the base
environment:
::
source activate base
Next, one can create the exoctk
conda
environment via the environment-3.6.yml
file:
(Both 3.6 and 3.7 are supported environments, but for our demo we encourage you
to use 3.6.)
::
conda env create -f environment-3.6.yml
Lastly, one can activate the newly-created environment and install the exoctk package with:
::
source activate exoctk-3.6 cd exoctk python setup.py [install|devlop]
As a final test, cd into your home directory, open up a quick connection to the
Python interpreter, and test that you can import exoctk
:
::
cd
python
>>> import exoctk