Skip to content

Latest commit

 

History

History
52 lines (40 loc) · 1.88 KB

INSTALL.md

File metadata and controls

52 lines (40 loc) · 1.88 KB
Raw

Installation Notes

The Monitor uses the LSST DM Stack's Butler, and also the SNCosmo package (among other things). Here's how to get everything working.

1. Get Anaconda Python

According to the LSST community, and various peopls around the DESC, it seems best to install the DM stack using conda - or in fact, miniconda. Click here and download python 2.7, if you don't have Anaconda python already.

bash ~/Downloads/Miniconda-latest-MacOSX-x86_64.sh

This installs miniconda into ${HOME}/miniconda2. You should now pre-pend your PATH with ${HOME}/miniconda2/bin so that this is becomes your default version of python.

2. Install the LSST DM Stack

First you need to add the LSST "channel:"

conda config --add channels http://research.majuric.org/conda/stable

Now do:

conda install lsst-distrib

and find something else to do for half an hour. This will install the LSST packages in ${HOME}/miniconda2/pkgs/.

3. Get set up to use the LSST DM Stack

See the project's notes here. LSST packages have to be added to your PATH, etc. before you can use them. This is handled by eups. For analyzing simulated images, we need to do:

    export LSST_DIR = ${HOME}/miniconda2
    set path = (${LSST_DIR}/bin $path)
    source ${LSST_DIR}/bin/eups-setups.sh
    setup obs_lsstSim

This needs to be done every time you start a new shell - so these lines could be worth adding to your .bashrc file or equivalent. Note that c-shell users can use eups_setup.csh and use setenv in their .login file.

4. Install python packages

We use several useful python packages that don't come with the DM stack:

pip install sncosmo
pip install pandas
pip install nose
pip install scipy