Artifacts

Data and code

Data, code, and instrument records from the research program. The current records are from the TESS transit-timing project; datasets, instrument designs, and calibration records from the other areas will be added as they are released.

Provenance policy

  • Every artifact has a repository path and, where a byte check applies, a SHA-256 digest.
  • A synthetic input is labelled synthetic in its own record. It is not an observation.
  • A real input names the archive, the request time, the HTTP status, and the local file.
  • If a number cannot be reproduced from the committed files, it is a draft, not a result.

Real data

The real material is public archive metadata and two SPOC product sets, for HD 209458 b (TIC 420814525), sector 56, and KELT-11 b (TIC 55092869), sector 9. No private or proprietary data is used. The archive query ran at 2026-10-22T06:15:21Z against the NASA Exoplanet Archive TAP service and the MAST Mast.Caom.Filtered endpoint.

Real data records and their repository paths.
Record Path What it contains
Archive run data/fetch-run/TARGETS.json Request times, HTTP statuses, response digests, and every evaluated row.
Target docket data/target-docket/docket.json 304 docket rows across 268 unique TICs, with one-sided sector status.
HD 209458 b acquisition data/real-run/acquisition.json Sector 56: source URLs, request times, sizes, and digests.
KELT-11 b acquisition data/real-run-2/acquisition.json Sector 9: source URLs, request times, sizes, and digests.
HD 209458 b single-transit fit data/sector56-control-run/analysis.json Mid-transit time, formal sigma, depth, duration, and residual RMS with the published period fixed.
KELT-11 b single-transit fit data/second-target-run/analysis.json Mid-transit time, formal sigma, depth, duration, and residual RMS with the published period fixed.
HD 209458 b full-sector fit data/sector56-full-run/run.json A full-sector reduction and fit on the same committed products.
Cross-check data/sector56-full-run/comparison.json Full-sector versus single-transit t0, depth, and RMS comparison. The paths disagree at the formal precision.
Light-curve diagnostics data/real-lc-run/diagnostics.json Cadence accounting, gaps, and scatter for the committed HD light curve.
Window QC data/real-window-qc/windows.json Coverage, background, and centroid summaries for the predicted windows.
Aperture comparison data/real-aperture-run/real_aperture.json Threshold aperture against SAP and PDCSAP on one common quality mask.

Synthetic fixtures

The photometric tests use local synthetic FITS and CSV files. These are test data, not observations, and not evidence about any star.

Synthetic inputs and their records.
Input Record
synthetic_transit.csv (27 d, 19,441 cadences) data/transit-fit-run/transit_fit.json
Injection campaign (12 synthetic light curves) data/injection-run/injection.json
Per-transit epoch measurements data/epoch-fold-run/epoch_measurements.json
Loader fixture and aperture fixture data/lightcurve-run/synthetic_lightcurve.json, data/aperture-run/synthetic_tpf.json

Reproducibility check

The committed artifacts are listed in data/repro-manifest.json. The current check reports 38 PASS and 0 FAIL against that manifest. The injection campaign is digest-checked rather than re-run because it takes about 160 seconds.

python research/tess-timing/scripts/verify_artifacts.py --check
python -m pytest research/tess-timing/tests -q

Code

The package is under research/tess-timing/. It reads archive metadata, applies the selection rule, loads light curves, performs aperture photometry, fits a transit model, measures per-transit times, and runs the injection campaign. The scripts that generate the committed records and the tests are described in the package README.

Licence status. No LICENSE file is committed at this revision. The code is readable but it is not offered under an open licence. Do not reuse it until a licence is chosen.

Figures

Every figure on this site is generated by a script in the repository. The captions name the generator. The four synthetic figures are the methods figure, the toy transit fit, the epoch fold, and the injection recovery.