CoolFace
Datasetpublic

kshitijd/mmu-l2-sdss

L2 model-ready views (release v1) The mmu-l2-* repositories provide processed, model-ready versions of the matching mmu-norm-* L1 data, while L1 keeps the normalized source measurements. L2 applies documented processing steps for training and evaluation, with the details for reversing each transformation stored in the row or in provenance.json. Repo View Reversal mmu-l2-tess per-sector relative flux f/median−1, time from first valid cadence flux =… See the full description on the dataset page: https://huggingface.co/datasets/kshitijd/mmu-l2-sdss.

sourceHugging Faceupdated 1mo agoView on Hugging Face
0likes106downloads
Dataset Card

L2 model-ready views (release v1)

<!-- Use this card (adjusting the first line) for: mmu-l2-tess, mmu-l2-sdss, mmu-l2-chandra, mmu-l2-sne, mmu-l2-provabgs -->

The mmu-l2-* repositories provide processed, model-ready versions of the matching mmu-norm-* L1 data, while L1 keeps the normalized source measurements. L2 applies documented processing steps for training and evaluation, with the details for reversing each transformation stored in the row or in provenance.json.

RepoViewReversal
mmu-l2-tessper-sector relative flux f/median−1, time from first valid cadenceflux = (rel+1)×flux_median_e_per_s; t = t_rel + t0
mmu-l2-sdssspectra zero-padded to 4,800 samples with valid maskdrop valid=false samples
mmu-l2-chandrabins zero-padded to 512 with valid maskdrop padding
mmu-l2-snetime-sorted sequences, band/survey tokens, ONE scale per object (flux: ÷ median \f\; mag: − median m), representation columnflux rows: v×scale; mag rows: v+scale
mmu-l2-provabgs16/50/84 posterior percentiles per parametersamples remain in L1

Each row keeps its assignment from splits/v1 (train, val, test, or unassigned). Images, DESI and VIPERS spectra, and Gaia already have a fixed shape, so their L2 preparation uses the L1 tensors with the training-only robust scaling statistics in mmu-norm-index under l2_stats/v1/ (z = (x − median) / iqr).

In every L2 view, padding is marked valid=false, one amplitude scale is used per object to preserve colors, statistics are fit on the training split only, and gaps are left as gaps.