SDSS-V BOSS spectra of the Culpan et al. (2022) sample, classified here and fitted against one grid.
2368 objects carry a class: 466 published, 2368 assigned by the rule-based classifier, 466 of them both. Judged with spclass.compare rather than by string — so that sdB against sdB: counts as agreement and sd against sdB as a refinement — the two are compatible for 308 of 466 and flatly contradict each other for 137. Those are what the review page is for. 428 have a profile fit in this tree and 87 of those a two-component SED analysis.
Counted on the best class each star has — published where there is one, the classifier's otherwise. A composite is counted under its primary here, so the 183 systems with a cool companion are spread across the rows above rather than forming one of their own.
Nothing on this page is a catalogue. Each entry below is one, and each star has its own page under its J-name.
Every star with a profile fit here, and the comparison that motivated them: the Bamberg v25 grid against Uli's on the same spectra.
The systems where the two-component SED fit itself keeps a cool companion: it improves the fit, its temperature is inside the grid, and its radius is a main-sequence radius.
Four specific claims, each judged on its own evidence. Their companions are cooler than the 3500 K grid floor, so the infrared leaves them unsettled rather than refuted.
Everything the published class, the classifier or the survey notes call a composite and the photometry has not confirmed.
Uli's grid against Bamberg's, star by star: what changes, and by how much.
Where the classifier and the published class disagree, and the spectra that would decide it.
To come. Each is one entry in bossindex.py: a file, a title, a sentence and a count.