astronomy-commons / astronomy-commons/data.lsdb.io
Import SDSS DR17 optical spectra
- Dominant language
- Jupyter Notebook
- Stars
- 1
- Forks
- 0
- Avg merge
- 21h 11m
- Merged PRs (30d)
- 4
Description
SDSS DR17 was the latest data release, which accumulated all up-to-date SDSS data, from SDSS I to SDSS IV. 18th and 19th DRs are focused on SDSS V only.
For now, let's scope on optical spectra, not infrared and not photometry. Here is the bulk-download documentation:
https://www.sdss4.org/dr17/data_access/bulk/
I also believe we don't need single-epoch spectra (until we'd like to practice in temporal-spectral data), so we can use `speclite` files, located here:
https://data.sdss.org/sas/dr17/eboss/spectro/redux/v5_13_2/spectra/lite/
FITS file schema:
https://data.sdss.org/datamodel/files/BOSS_SPECTRO_REDUX/RUN2D/spectra/lite/PLATE4/spec.html
Fields I propose to use:
- **HDU 0 Header**: nothing
- **HDU 1 Data** - spectrum: we take everything as a single nested column (variable length list), everything but "model" should be default sub-columns
- **HDU 2 Data** - specObj: I believe this is always a single-item list (better to have an assert to check). Let's take all the columns here (I have no idea if there are any useless stuff there). Special columns:
- PLUG_RA, PLUG_DEC - coordinate columns
- OBJID - it a 5-tuple for the object ID, let's split into OBJID_RUN, OBJID_RERUN, OBJID_CAMCOL, OBJID_FIELD, OBJID_ID
- All other 5-tuples (float32[5] type) are related to 5-band (ugriz) photometry, I propose to split them into individual columns with "_{band}" suffixes
- Let's make all 10-tuples to be fixed-length-list columns
- Default columns: SPECOBJID, Z, Z_ERR, CLASS, CLASS_ERR
- **HDU 3 Data** - spectral lines: we take everything but "PLATE", "MJD" and "FIBERID" (duplicated from HDU 2) as a single nested column (I believe it may be a fixed-length-list column, length is 32). No default columns here.
Catalog name would be something like "sdss_dr17_eboss"
Contributor guide
No contributing guide indexed for this repository
Research direction
Start with the SDSS DR17 bulk-download documentation and the FITS file schema linked in the issue, then inspect the speclite files under the provided data path. Map HDUs 1–3 to the proposed nested, fixed-length, coordinate, identifier, and photometry columns. Done means an sdss_dr17_eboss catalog imports optical spectra with the specified exclusions and field transformations.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- jupyter-notebook
- Domain
- data
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Quiet
- Clarity
- Mostly clear
- Newbie friendliness
- 48/100