Skip to content

RichardsGroup/LSST_OpSim

Repository files navigation

LSST_OpSim

The repo provides instructions, notebooks, and scripts to help people getting started with testing LSST cadence simulations hosted on SciServer. This work was initiated to encourage LSST AGN SC members to design and run metrics on simulated LSST cadences, to enable selection of cadences that are best for AGN science in LSST. However, since the data and notebooks are all public, anyone interested in testing the cadences is welcome to use the tools provided here within the infrastructure provided by the SciServer team. This project is led by Weixiang Yu and Dr. Gordon Richards at Drexel University, with extensive help from the SciServer team.

Setup

The instructions on how to create an account on SciServer, create a container with the necessary software installed and the cadence volume mounted can be found in sciserver_opsim.pdf

Installing rubin_sim

This section is added following the release of a new rubin_sim Python package for running MAF on simulated LSST surveys and performing other survey simulation tasks. rubin_sim replaces the old tools that require the LSST Stack to be functional. Note that the instructions provided here are for running rubin_sim on SciServer only. A more general-purpose installation instruction of rubin_sim is available at the original rubin_sim github repository. You can follow the steps below to install rubin_sim.

  1. Navigate to your persistent folder at '/home/idies/workspace/Storage/{your_username}/persistent' from the terminal.
  2. Clone the rubin_sim repository and install from source:
    git clone https://github.com/lsst/rubin_sim.git
    cd rubin_sim
    conda create -n rubin -y
    conda activate rubin
    conda install -c conda-forge --file=requirements.txt -y
    pip install e .
    ln -nsf ~/workspace/lsst_cadence/fbs2/ ~/rubin_sim_data # tell rubin_sim where to find data
  1. Install and create an ipython kernel for the rubin conda environment
    conda install ipykernel -y
    python -m ipykernel install --user --name rubin --display-name "rubin"

From now on, the rubin_sim package should be accessible in both a script and a notebook (you need to select the "rubin" kernel in the notebook interface).

Note:

If a link of ~/rubin_sim_data already exists, please remove it and create a new one.

Getting Started

Once you have finished the setup, clone this repo to your persistent folder. You can begin exploring the simulated cadences using the notebooks and scripts provided here. We would suggest following the order listed below:

  • Introduction.ipynb: A notebook providing a brief overview about how to use MAF.
  • Multiple_Opsims.ipynb: A notebook showing how to run some metrics on multiple (all) opsims.
  • View_Results.ipynb: A notebook showing how to read in the result produced in the notebook above.
  • wfdFootPrint.ipynb: A notebook showing how to use a custom healpix slicer to run metrics on WFD observations only. Since the Feature-based opsims no longer use fixed tiles, we have get WFD observations through some tricks. For more discussions on this topic, please see this thread on LSST community.com
  • DDF_Other_FootPrint.ipynb: A notebook showing how to run metrics on DDF only or areas that are outside DDF and WFD.
  • rubin_sim_notebooks: A collection of MAF tutorial notebooks provided by the Rubin Project Team. These notebooks have been modified slightly to accommodate the latest rubin_sim API.

Note: The opsimUtils.py script must be kept in the same directory in which you want to run the notebooks. This file is also constantly updated to match the schema changes in the newer opsims.

Useful resources:

  • A summary/cheetsheet (by Lynne Jones) of cadence simulations that are currently available.
  • Columns in the OpSim database -> here
  • A high-level comparison/description of various simulation groups/families (across FBS 1.5, 1.6 and 1.7)-> here

Pro tip: If you have already walked through all of the notebooks provided above and realize that some code might take forever to run, you can experiment with the SciServer Jobs (which can give you the access to more computing power and memory), here is how to do it -> SciServer_Jobs.

About

Notebooks and tools to perform MAF on OpSims

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published