Run on Rubin+Roman data on RomanDESCSim

[wmwv]

Run a supernova from RomanDESCSim with both Roman and Rubin/LSST data

1. One SN
2. Ten SNe

[wmwv]

The Rubin/LSST images haven't been processed. Need some way of getting the image list. If we use the eimage files and assume the WCS is perfect we might be able to do this still.

[wmwv]

On NERSC, the RomanDESCSim data are in

/global/cfs/cdirs/lsst/production/roman-desc-sims/

With RomanWFI/TDS data in

/global/cfs/cdirs/lsst/production/roman-desc-sims/Roman_data/RomanTDS/images/simple_model

And Rubin/LSST data in

/global/cfs/cdirs/lsst/production/roman-desc-sims/raw_data/

There's a sample reduction for the Rubin/LSST data done (by Jim Chiang?) with repo and collection

repo = "/global/cfs/cdirs/lsst/production/roman-desc-sims/preview_data_staging/preview_data/"
collection = "u/descdm/preview_data_step1_w_2024_12/20240326T152819Z/"

[wmwv]

Ah, I was confused. Correct repo, collection

repo = "/global/cfs/cdirs/lsst/production/gen3/roman-desc-sims/repo"
collections = ["u/descdm/preview_data_step1_w_2024_12"]

[wmwv]

Here's a snippet to get the matching images for a given RA, Dec

from lsst.daf.butler import Butler
from lsst import sphgeom

transient_id = 30328322
ra, dec = 8.52941151,-43.0266337
mjd_start, mjd_end = 62300.0, 62600.0

repo = "/global/cfs/cdirs/lsst/production/gen3/roman-desc-sims/repo"
collections = ["u/descdm/preview_data_step1_w_2024_12"]

# The Step3/coadd collection is:
# collection = ["u/descdm/preview_data_step3_2877_19_w_2024_12"]

butler = Butler(repo, collections=collections)
butler.registry.queryCollections()

level = 10  # the resolution of the HTM grid
pixelization = sphgeom.HtmPixelization(level)

htm_id = pixelization.index(
    sphgeom.UnitVector3d(
        sphgeom.LonLat.fromDegrees(ra, dec)
    )
)

dataset_refs = butler.registry.queryDatasets("calexp", htm20=htm_id)

[wmwv]

Reading and processing Roman and Rubin data !

[wmwv]

Next need to rethinking dataset, which is now per-function, but needs to be per row.

[wmwv]

Dropped dataset. Just tracking based on instrument.

[wmwv]

Next steps:

1. [ ] Read truth file for region, (ra, dec)=(8.5, -43) with nside=32
2. [ ] Pick some SN w/ roman+Rubin time (and space) overlap. From set of already processed Rubin data (from RomanDESCSims)
3. [ ] Plot sampled model LCs from truth files for selected SN (both new and exist). Save LC data files and plots
4. [ ] Construct transient info and transient info host/scence tables
5. [ ] Get image list
6. [ ] Run

[wmwv]

Selecting supernova with:

import numpy as np
from astropy.table import Table

file = "/global/cfs/cdirs/descssim/imSim/skyCatalogs_v1.1.2/snana_10306.parquet"
df = Table.read(file)
w, = np.where((df["z_CMB"] < 0.7) & (df["ra"] > 8.3) & (df["ra"] < 8.7) & (df["dec"] > -43.5) & (df["dec"] < -42.5) & (df["start_mjd"] > 62200))

yields 391 candidates.

[Edit: Give absolute path for skyCatalogs.]

[wmwv]

Of the first 10 and last 10 candidates of those 391, the following three had Rubin simulated images during the SN that have been processed out of the RomanDESCSims:

50130277 50132692 110000220

had matching images.

[wmwv]

Updated to include collection with many more (all?) of the images. Getting much long before and during lists now. (Not as many after, because the end date is super conservative and the survey ends).

[wmwv]

import numpy as np
from astropy.table import Table

file = "/global/cfs/cdirs/descssim/imSim/skyCatalogs_v1.1.2/snana_10306.parquet"
df = Table.read(file)
w, = np.where((df["z_CMB"] < 0.4) & (df["ra"] > 8.3) & (df["ra"] < 8.7) & (df["dec"] > -43.5) & (df["dec"] < -42.5) & (df["start_mjd"] > 62200))
df[w]["id", "ra", "dec", "z_CMB", "start_mjd", "end_mjd"].write("list.csv")