astro-legacy-archive/nustar-fpma-cleaned-events

NuSTAR FPMA Level-2 events for 3C 273

[image]

This dataset contains the NuSTAR Level-2 cleaned science-quality FPMA event file nu10012001002A01_cl.evt.gz for observation 10012001002, a calibration and cross-calibration pointing at the quasar 3C 273 (OBJECT=3C273_pcs21). A identifies FPMA and 01 the science-quality observing mode. The source is `nu10012001002A01_cl.evt.gz`, checked 2026-08-30: 4,507,983 bytes, SHA-256 930e077de95790ef1c37d5abe087fb6c3566ece7431b881f6e192a882da9204b.

python -m venv .venv && .venv/bin/pip install datasets huggingface_hub pyarrow

from datasets import load_dataset
dataset = load_dataset("astro-legacy-archive/nustar-fpma-cleaned-events", "nu10012001002A01_cl__EVENTS", split="train[:1]")
print(dataset.num_rows, dataset[0]["TIME"], dataset[0]["PI"])

1 79840651.76565212 11

import json
from huggingface_hub import hf_hub_download
import pyarrow.parquet as pq
path = hf_hub_download("astro-legacy-archive/nustar-fpma-cleaned-events", "nu10012001002A01_cl__EVENTS.parquet", repo_type="dataset")
print(json.loads(pq.read_metadata(path).metadata[b"event_list"])["event_count"])

21997

Source structure and decoding

The published table is the complete FPMA-field EVENTS binary table: 21,997 rows and these 39 source columns in source order:

TIME 1D s; PRIOR 1D s; NUMRISE 1J NA; DENRISE 1J NA; SUBFRM 1J NA; PREPHAS 9I NA; POSTPHAS 9I NA; RESET 1D s; TRIGGERS 9X NA; DET_ID 1B NA; S_CAP 1B NA; HWTRIG 1B NA; RAWX 1B NA; RAWY 1B NA; MODE 1B NA; STIM 1B NA; UP 1B NA; SHIELD 1B NA; SHLD_T 1B NA; SHLD_HI 1B NA; STATUS 16X; BADPOS 8X; HOTPOS 8X; RAWPHAS 9I; OFFPHAS 9E; TRPHAS 9E; PHAS 9E; GRADE 1I; SWTRIG 9B; PIS_GAIN 9E chan; SURRPI 1J; PI_CLC 1E; PI 1J chan; DET1X 1I pixel; DET1Y 1I pixel; DET2X 1I pixel; DET2Y 1I pixel; X 1I pixel; Y 1I pixel

The exact meaning and representation of every column resolve through the embedded event_list document: iterate metadata["fields"].items() in order. Each entry gives description, unit, source dtype, TTYPE, TFORM, TUNIT, scaling, null, display and dimension declarations, served encoding, repeat, bit mapping and any special values. fits_sources[0].fits_headers carries the complete ordered primary and extension headers, including the source comments; auxiliary_tables gives every companion definition and every value; sources gives the URL, size, hash and check date. Thus a consumer can write a resolver from the Parquet file alone. load_dataset retains served values and types but not this schema metadata; read it with PyArrow as above.

The vectors PREPHAS, POSTPHAS, RAWPHAS, OFFPHAS, TRPHAS, PHAS, SWTRIG and PIS_GAIN retain nine elements per row. TRIGGERS is a nine-bit field; STATUS is 16 bits; BADPOS and HOTPOS are eight bits. Each is served in stored FITS element order without repacking. RAWX/RAWY are physical pixels within DET_ID; DET1X/DET1Y are fixed focal-plane coordinates; DET2X/DET2Y apply the time-dependent mast transformation; X/Y are FK5 J2000 TAN sky pixels.

Five complete companion tables are embedded in HDU order. GTI has seven rows of START 1D s; STOP 1D s. Four distinct extensions retain their repeated EXTNAME=BADPIX and source order; each has RAWX 1B pixel; RAWY 1B pixel; TIME 1D s; TIME_STOP 1D s; BADFLAG 16X: EXTVER=1 / DETNAM=DET0 has 7 rows, 2 / DET1 has 28, 3 / DET2 has 15 and 4 / DET3 has 28.

TIME ranges from 79840651.76565212 to 79849481.7151812 s and every value is distinct. PI ranges from 0 to 4095. Its nominal energy mapping is 1.6+0.04*PI keV; NuSTAR's normal science band is approximately 3--79 keV, and spectroscopy requires the region-matched RMF and ARF. PI_CLC contains 928 source NaNs, including 907 at GRADE=0; they are intentional uncomputed charge-loss values, not Arrow nulls.

Scientific caveats

Level 2 means calibrated and screened, not analysis-ready or source-extracted. This is the complete FPMA field, not a source-region event sample, and the separately served FPMB product is required for the other half of the observation. Select source and background regions, build module- and region-specific RMF, ARF and exposure products, and analyse FPMA and FPMB separately or jointly.

The source records TIMESYS=TT, TIMEREF=LOCAL, CLOCKAPP=F, TIMEZERO=0, MJDREFI=55197 and MJDREFF=0.00076601852. ONTIME=5609.775005578995 s, LIVETIME=EXPOSURE=5148.77021417426 s and DEADC=0.917821161999142; the event interval spans 8831.25 s. These times are spacecraft-local, not barycentric. Absolute or source-frame timing requires the current NuSTAR clock-correction file, attitude/orbit product and barycorr.

The file embeds NuSTARDAS 1.9.2 / HEASoft 6.27.1 and CALDB 20200429. That frozen processing predates current calibration and clock products. HEASARC's current FPM CALDB is 20260807, released 2026-08-10 and including clock-correction file v231; current analysis guidance is to rerun nupipeline from archive inputs with current NuSTARDAS and that CALDB.

The archive directory also serves the corresponding FPMB science event file, other observing-mode files, separate GTIs, images, regions, mast, attitude, orbit, detector-coordinate and housekeeping products. Observation 10012001002 is the 2012 3C 273 calibration example in the NuSTAR Quickstart Guide and part of the multi-observatory cross-calibration campaign. The Parquet artifact has 21,997 rows and 6,046,490 bytes, SHA-256 6f43c21718f42e59bdc1f7c1e27c6e93e392f3b5bc6d1a5011869aff997a56fa.

Rights, credit and citations

No explicit data licence was found. HEASARC states:

HEASARC materials are all available freely for your use.

HEASARC requests:

This research has made use of data and/or software provided by the High Energy Astrophysics Science Archive Research Center (HEASARC), which is a service of the Astrophysics Science Division at NASA/GSFC.

Archive and analysis references: NuSTAR archive products, Quickstart Guide, FAQ and CALDB index. Please cite Harrison, F. A. et al. (2013), ApJ 770, 103, DOI 10.1088/0004-637X/770/2/103, arXiv:1301.7307, and Madsen, K. K. et al. (2017), AJ 153, 2, DOI 10.3847/1538-3881/153/1/2, arXiv:1609.09032.

Errata and requests

Report corrections or dataset requests in the dataset's Community tab.