Instruction for Tutorials

Two tutorials are scheduled for an hour and a half each. To be more effective it is necessary to do some preparation work - install software etc.

Instructions for necessary preparatory steps for Tutorial:The LSST SOFTWARE STACK.

This tutorial will walk participants through writing a simple script to process detrended image files to generate catalogs using the DM stack, and then demonstrate some techniques for using the stack to analyze catalogs and images.  If time permits, we may briefly walk through the steps involved in running the current production pipeline prototype on data from an observatory the stack has already be specialized to work with.  Participants who want to follow along *must* install the stack and download a small amount of example data in advance.
 

Installation instructions can be found here (conda binaries are the easiest approach for most platforms):
https://pipelines.lsst.io/install/index.html

Example data downloads can be found here:

 
We recommend that anyone interested in participating in the tutorial download the smaller file, example1.fits (57 MB).  The larger tarball (2.2GB) is optional, but may be useful for anyone looking for in-depth help running the pipeline on real data.
 
 

Instructions for necessary preparatory steps for Tutorial:LSST Simulations

The accurate detection and characterization of variable and transient sources is an important science goal for the LSST.  In order to evaluate  the performance of any proposed observing cadence at meeting that goal, researchers must have access to realistic simulations both of the variable/transient population of the universe and of the behavior of LSST itself.  OpSim (the Operations Simulator) is a software tool designed by the LSST project to take fiducial scheduling algorithms and produce high fidelity simulations of the 10 year observing cadences they produce.  OpSim includes accurate characterizations of all of the expected mechanical performances and tolerances of the LSST telescope and is designed to interface with scheduling algorithms exactly as the actual telescope will. CatSim (the Catalog Simulator) is a series of data and software products designed to provide users with an accurate simulated fiducial universe to observe with the OpSim simulated survey.  It contains distributions of galaxies drawn from N-body simulations, distributions of Milky Way stars based on SDSS data, and SED and variable light curve libraries based on the latest theoretical and phenomenological models.  In this tutorial, we will demonstrate how to use OpSim and CatSim to produce accurate LSST light curves for variable and transient populations with photometric uncertainties expected from the LSST system.  This should allow researchers to effectively evaluate the performance of LSST at identifying and characterizing variable populations of their choice.

 

Before attending this tutorial, participants will want to install CatSim and download a fiducial OpSim output.

To install CatSim (please do not do this until 3 days before the conference, we are still optimizing some of the functionality that will be demonstrated):

1) Install Miniconda.  Miniconda is a software distribution tool (like pip or yum) used to distribute python and many of the most useful python-based software tools.  If you do not already have Miniconda (or the heavier-duty Anaconda), go to: http://conda.pydata.org/miniconda/html and follow the installation instructions found there.  Be sure to install the python 2.7 version.  The LSST Simulations stack is not yet python 3 compliant.

2) Miniconda provides you with a distribution of python as well as the software tool conda for downloading and installing new packages.  In order to use these tools, they must be on your PATH.  Prepend the directory $MINICONDA_HOME/bin to your PATH (where MINICONDA_HOME is the directory where Miniconda was installed).

If you are running a bash shell, this will look something like:

    cd my/miniconda/directory/
    export PATH=$PWD/miniconda2/bin/:$PATH

you should now be able to type

    which conda

and see the path to the conda executable installed by Miniconda

3) Miniconda's conda tool installs software from a series of distribution servers that Miniconda knows about.  Now you must tell Miniconda about the server that contains the LSST Simulations software.  Run the command

    conda config --add channels http://eupsforge.net/conda/dev

4) Install MAF by running

    conda install lsst-sims

This will take about half an hour.

5) Confirm that the installation was successful.  There are a few steps you will need to run through whenever you want to use CatSim.

    a) Open a new terminal.
    b) Add $MINICONDA_HOME/bin to your PATH as in step(2)
    c) Run

          source eups-setups.sh

       This sets up a version of EUPS, the package management software that handles version control for LSST Software.

    d) Run

           setup lsst_sims -t conda

       This will tell EUPS to setup the conda-installed version of CatSim.

    e) Now, verify that MAF has successfully been installed by opening a python session and try

           import lsst.sims.utils

       If you do not get any errors: congratulations! You have successfully installed CatSim. We will show you want to do with it in Belgrade.

To download the OpSim data, go to

    http://lsst.org/scientists/simulations/opsim/opsim-v335-benchmark-surveys

scroll down to the table at the bottom of the page, and download the SQLite Data for kraken_1042.  This will be about 5GB of data.