astro-legacy-archive/ixpe-level2-events

IXPE Level-2 events

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

This dataset contains the 515,004-row detector-unit-3 Level-2 field event table for aggregate observation 01003199. The FITS header records OBJECT='1RXS J170849.0'. These are all DU3 field events, not an extracted or background-subtracted source sample. DU3 is one of three co-aligned detector units; scientific products normally combine DU1, DU2 and DU3 using their corresponding responses.

from datasets import load_dataset
dataset = load_dataset("astro-legacy-archive/ixpe-level2-events", "ixpe01003199_det3_evt2_v01__EVENTS", split="train[:1]")
print(dataset[0]["TIME"])

180335311.57109

import json
from huggingface_hub import hf_hub_download
import pyarrow.parquet as pq
path = hf_hub_download("astro-legacy-archive/ixpe-level2-events", "ixpe01003199_det3_evt2_v01__EVENTS.parquet", repo_type="dataset")
print(json.loads(pq.read_metadata(path).metadata[b"event_list"])["schema"])

astro-legacy-archive/event-list

Essential provenance and interpretation

The archive served only v01 for this observation when checked on 2026-09-05. The pinned file's internal DATE=2022-10-12, PROCVER=00.02, and CALDBVER=20211118 predate pipeline instrumental-background flagging: STATUS is all zero, no XPFLGBGD header keyword is present, and instrumental background remains unflagged. Newer products carry the event-level background flag in STATUS bit 8. Apply the current IXPE pre-flagging background workflow plus energy, spatial/source, and STATUS2 selection before treating rows as a source ensemble. For this aggregate DU3 file, filter_background.py uses the corresponding Level-1 files ixpe01003101_det3_evt1_v02.fits.gz, ixpe01003102_det3_evt1_v03.fits.gz, and ixpe01003103_det3_evt1_v02.fits.gz. At filter_background.py revision 71719cb, captured 2026-09-06, path_lv2 accepts one Level-2 file and path_lv1 uses nargs='+'; the script appends the event arrays from every supplied Level-1 file. The interface for this aggregate file is:

python filter_background.py ixpe01003199_det3_evt2_v01.fits.gz ixpe01003101_det3_evt1_v02.fits.gz ixpe01003102_det3_evt1_v03.fits.gz ixpe01003103_det3_evt1_v02.fits.gz

The pinned upstream FITS has invalid internal integrity cards. The primary, EVENTS, and GTI CHECKSUM values are respectively X4aaY2RWX2XaX2XW, XBAra93oRAAoX73o, and UPBbaP9aWPAaaP9a; calculation from the pinned file gives 2TaW3RWV2RaV2RUV, JAgAK2Z2J9d9J9Z9, and eM6RfM6PeM6PeM6P. The primary DATASUM=0 is valid. The stored EVENTS and GTI values 2053291623 and 3586949736 are invalid; the computed values are 3880052192 and 385960213. These source cards remain unchanged. The external SHA-256 below identifies the compressed source bytes. event_list.verify() checks that pin, stored headers, represented table structure and values, and reconstruction; it does not require the source FITS CHECKSUM or DATASUM assertions to be valid. The same results are machine-readable under source_integrity in the Parquet metadata.

TRG_ID is the trigger identifier; the source distinguishes triggers because the data-acquisition system can discard them according to region-of-interest size. TIME is the sum of the source seconds and microseconds in IXPE Time, measured in TT seconds from MJDREFI=57754 and MJDREFF=0.00080074074074. X and Y are Cartesian pixels on the J2000 TAN plane, not longitude and latitude. Their source WCS uses TCTYP7=RA---TAN, TCTYP8=DEC--TAN, reference pixel (299, 299), reference coordinate (257.19529, -40.1479) degrees, and increments (-0.00072222, 0.00072222) degrees per pixel.

PI maps nominally to energy as E = 0.02 + 0.04*PI keV. The file spans PI 1--374 (nominally 0.06--14.98 keV), but the recommended operational band is 2--8 keV; the current guide describes the upper response limit as about 10 keV. Q and U are event pseudo-Stokes contributions, and W_MOM is their moments-analysis statistical weight. Their values—including Q/U outside ±1—are not per-photon polarization. Combine them only over selected event ensembles using a documented weighting convention and matching RMF/ARF/MRF products.

For HEASoft 6.35 extractor polarization spectra, the canonical IXPE analysis page says the “normalization of Q and U in resultant spectral files are incorrect.” It also says “HEASoft 6.35.1 contains an update to the extractor task to fix this issue.” Use 6.35.1 or later for those spectral extractions; see Known IXPE Analysis Issues.

STATUS2 is set in 519 events: 226 carry surviving Level-2 bit 1 (invalid GPS position) and 293 carry bit 12 (inside the SAA boundary), with no overlap or other set bit. The source GTI table has 761 intervals. Header ONTIME=837863.027258 s is engineering-defined exposure minus missing telemetry intervals. The Level-2 intervals satisfying the pointing, filter-wheel, and detector-mode criteria sum to 837842.7306529582 s, a 20.296605 s difference. The pinned HEASARC IXMASTER result for select * from ixmaster where obsid='01003199' records ontime_3=5772.014 s and exposure_3=5766.4362 s; IXMASTER documents these as DU3 on-source time and DU3 effective exposure on source. The pinned served FITS EVENTS header records ONTIME=837863.0272580001 s. These HEASARC surfaces disagree. The GTI values are stored under event_list.auxiliary_tables; all three source-header records are under event_list.fits_sources[].fits_headers. Times are TT/LOCAL and are not barycentric.

Observation 01003199 aggregates segments 01003101, 01003102, and 01003103. Their attitude products are required for observation-specific response work. For DU3 these are ixpe01003101_det3_att_v03.fits.gz, ixpe01003102_det3_att_v01.fits.gz, and ixpe01003103_det3_att_v01.fits.gz. The aggregate event file identifies OBS_ID=01003199; it does not contain a segment column or segment-boundary mapping.

Source pin

Observed 2026-09-01: ixpe01003199_det3_evt2_v01.fits.gz, 19,594,230 bytes, SHA-256 29c3726a103cb36811010f113290d41d8b005c3ed7d18ab17b2168fa7ef9b196.

Cite Weisskopf et al. (2022), DOI 10.1117/1.JATIS.8.2.026002. No explicit dataset licence was found. HEASARC's free-use statement and requested acknowledgement apply.

IXPE strongly recommends the following mission-specific acknowledgement:

This work reports observations obtained with the Imaging X-ray Polarimetry Explorer (IXPE), a joint US (NASA) and Italian (ASI) mission, led by Marshall Space Flight Center (MSFC). The research uses data products provided by the IXPE Science Operations Center (MSFC), using algorithms developed by the IXPE Collaboration (MSFC, Istituto Nazionale di Astrofisica - INAF, Istituto Nazionale di Fisica Nucleare - INFN, ASI Space Science Data Center - SSDC), and distributed by the High-Energy Astrophysics Science Archive Research Center (HEASARC).

HEASARC separately requests its service acknowledgement:

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.

Citation

Please cite this dataset:

@misc{https://doi.org/10.57967/hf/10405,
  doi = {10.57967/hf/10405},
  url = {https://huggingface.co/datasets/astro-legacy-archive/ixpe-level2-events},
  author = {{Astro Legacy Archive}},
  title = {ixpe-level2-events},
  publisher = {Hugging Face},
  year = {2026}
}

The initial public release is Hub revision 0746b24685b02c395fc14d54de24f89f8b04c559.

Errata and requests

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