Simulating the Roman High Latitude Time Domain Survey with SNANA

   ___                         _____  __ 
  / _ \___  __ _  ___ ____    / __/ |/ /
 / , _/ _ \/  ' \/ _ `/ _ \  _\ \/    /
/_/|_|\___/_/_/_/\_,_/_//_/ /___/_/|_/

These are the input files for SNANA/Pippin catalog-level simulations of the Roman High-Latitude Time Domain reference survey as presented in 2111.03081. They allow for the simulations of extra-galactic transients as may be observed by the Roman Space Telescope. These simulations allow for both the creation of a full transient catalog and a supernova cosmology pipeline.

Usage

To generate a full transient (>12 types) catalog, run

pippin.sh {your initials}-ROMAN-TRANS.yml

The supernova cosmology pipeline can be run via

pippin.sh {your initials}-ROMAN-COSMO.yml

I recommend that you read the Pippin best practices and follow those when running these simulations.

The intended use, for most people, is to analyses the resulting simulations via a data release, rather than modify the input files themselves themselves. Though, if you intend to look at variations in the simulations, you can use these input files wherever you have access to SNANA. I would recommend running on Midway (operated by U. Chicago), NERSC, or contacting one of the (contributors)[#Contributors] and asking them to run the simulations.

Requirements

Set up

Full set up:

Download via git clone https://github.com/benjaminrose/roman_snana_sims.git
Change the name of PIPPIN_ROMAN_TRANS.yml and PIPPIN_ROMAN_COSMO.yml files to avoid overwriting other simulations., e.g. PIPPIN_ROMAN_TRANS.yml --> {your initials}_ROMAN_TRANS.yml
Add an environment variable (HLTDS_SIM_BASE_PATH) that points to the location to your simulation input files. This variable needs to be in your startup script (i.e. .bashrc) so that batch jobs also see it.

You can do this with echo "export HLTDS_SIM_BASE_PATH=$(pwd)" >> ~/.bashrc

Check that your setup is correct with ... (pippin.sh -c, HLTDS_SIM_BASE_PATH exits, and file name has changed)
Check the pipeline files before launching simulations

pippin.sh -c {your initials}_ROMAN_TRANS.yml

In the future, we hope that you can point to a managed version of these files rather then cloning the entire repo.

Update to the latest sims

git pull
Again, rename PIPPIN_ROMAN_*.yml files

Support

Currently there should be a minimal expectation of support. In time, we hope to make this releasable with a full support structure. For now, these exist publicly more as a reference than for broad usage.

Project Structure

These files follow the SNANA pipeline nomenclature as defined by PIPPIN.

1_sim/ - simulating
- searcheff/roamn_deep+subaru -
- roman_ref.simlib - input to recreate Rose+ 2021
- sim_foundation.input - information regarding the Foundation survey and instruments.
- sim_roman.input - information regarding the Roman instrument/survey
- simgen_*.input - files about how each transient should be simulated (i.e., links to the SEDs, rate as a function of redshift, ...)
2_lcfit/ - Light-curve fitting
- global.yml - a list of systemics to test/change as a part of the light curve fitting and covariance steps
- SIMFIT_*.nml - survey sepcific LC fitting settings.
3_clas/ - files realted to photometric classification of transients
- models/ - contains the trained classifiers, one folder per classifier
- train/ - contains PIPPIN pipelines to train (and validate the training) of photometric classifiers to work with Roman data.
6_biascor/
- BBC_ROMAN.input - settings for the BBC step in the SN cosmology pipeline.
PIPPIN_ROMAN_TRANS.yml - Pippin pipeline file used to run these simulations.

Simulations files that are stored outside this folder

Filter and prism definitions
- /project2/rkessler/SURVEYS/ROMAN/ROOT/kcor/H19/kcor_WFIRST_CYCLE8_IMGPRISM_2021v3.fits
Selection
- /project2/rkessler/SURVEYS/ROMAN/ROOT/models/searcheff/SEARCHEFF_PIPELINE_WFIRST.DAT
Host galaxies
- /project2/rkessler/SURVEYS/ROMAN/ROOT/starterKits/sim_host_redshift/AKARI+specbasis_v1.5.1.HOSTLIB.gz
- (referenced but not used) /project2/rkessler/SURVEYS/ROMAN/USERS/pmacias/dev/sim/searcheff/dust/new/roman_deep+subaru
Transient Models
- /project2/rkessler/PRODUCTS/SNDATA_ROOT/models/NON1ASED/NON1ASED.P18_CC
- /project2/rkessler/SURVEYS/ROMAN/ROOT/SALT3-NIR_dev/SALT3.P22-NIR
- /project2/rkessler/SURVEYS/LSST/ROOT/PLASTICC/
  - model_libs/NON1ASED.TDE-MOSFIT
  - model_libs/NON1ASED.CART-MOSFIT
  - model_libs_updates/NON1ASED.V19_CC+HostXT
  - model_libs/NON1ASED.ILOT-MOSFIT
  - model_libs_updates/NON1ASED.BULLA-BNS-M2-2COMP
  - model_libs_updates/NON1ASED.SNIax
  - model_libs_updates/NON1ASED.SNIa-91bg
  - model_libs/NON1ASED.PISN-MOSFIT
  - model_libs/NON1ASED.TDE-MOSFIT
  - model_libs/NON1ASED.SLSN-I-MOSFIT
Definition of the Foundation survey's observations
- to be determined.

Converting SIMSED_BINARIES to NON1ASED

(Re-)Creating SIMSED_BINARIES

This is no longer needed now that we are using NON1ASED rather then binaries.

This needs to be redone with every update to the filters (the kcor file) or the simulation (the simlib file).

go to /project2/rkessler/SURVEYS/LSST/ROOT/PLASTICC/model_libs/SIMSED_BINARIES/config_NGRST
run make_simsed_binaries.py make_simsed_binaries_NGRST.yaml >& make_simsed_binaries_NGRST.LOG
check that all the new SIMSED binary files have been created
cp them one folder up

Contributors

Ben Rose
Maria Vincenzi
Rebekah Hounsell
Phil Macias
Rick Kessler

Known issues

low-z input file needs to not have redundant SNIa information (i.e. GENALPHA_SALT2, GENMODEL, ...)

benjaminrose / roman_snana_sims