Headers: document AST_CAT, PT_CAT and PH_CAT (issue #1347) The keywords name the astrometric and photometric reference catalogs actually used, added in AstroPack bdb635ca8. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Updated New functions and changes log (markdown)
LAST bit mask: note that CR_DeltaHT is a sharpness criterion, not a cosmic-ray identification (issue #1333) The bit is set when the delta-like template S/N exceeds the PSF template S/N, so hot/noisy pixels and marginal detections satisfy it as well as real cosmic rays. Records the measured numbers, how to get at a realistic CR rate, that the flagged sources are removed from the catalog, and that the bit propagates into the coadd with an OR. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Changes log: Sasha Krassilchtchikov, 2026 Sep 16/23/24 entries
SN_ITER1 column: Catalogs, MatchedSources and changes log (issue #1113)
Updated LAST_DP_PipeI_Header (markdown)
Destroyed The colour–spectral slope relation α(BP−RP) (markdown)
Updated LAST_DP_PhotometricCalibration (markdown)
Created The colour–spectral slope relation α(BP−RP) (markdown)
Updated _Footer (markdown)
Updated _Sidebar (markdown)
Created LAST_DP_PhotometricCalibration (markdown)
New functions log: io.tiff and ultrasat.lab (issue #1283)
New functions and changes log: Sasha Krassilchtchikov, Aug 24 - Sep 10 2026
LAST_DP_PipeI_Catalogs: fix the MathJax errors in the new chi-square section Markdown turns the escaped \_ of \texttt{PSF\_CHI2DOF} back into a bare _ before MathJax sees it, and \texttt puts its argument in text mode, where _ is illegal - so all four display equations rendered as "'_' allowed only in math mode". The catalog column is now written as (chi^2/dof)_cat, a symbol defined in the prose, and the column name stays outside the math in backticks. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UQWbMJJMKCBehGkD9ZFUbc
LAST_DP_PipeI_Catalogs: PSF-fit chi-square and the systematic-error floor (issue #1271) PSF_CHI2DOF is ~1 only for faint sources and grows as SN^2 for bright ones. The new section derives why - the least-squares flux absorbs the PSF-parallel part of the model error, so only the orthogonal (shape) part reaches chi^2, giving <chi2/dof> = 1 + (eps*SN)^2/dof - gives the corrected statistic and the SN-dependent rejection threshold, a recipe to re-measure eps and dof, the values measured on one visit (eps=0.073, dof=24.6, slope 2.12), and the result that eps is neither equal nor proportional to the photometric-calibration rms. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UQWbMJJMKCBehGkD9ZFUbc
log pipeline.last.quality.raw and removeRaw (issue #1228)