astro-legacy-archive/hitomi-sxs-cleaned-events

Hitomi SXS cleaned events

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

[image]

This dataset contains two archive-natural Hitomi SXS cleaned-event products as separate configurations. Observation 100040010 (Perseuscore) is the leading science configuration: the official observation log marks the source in the SXS field of view. Observation 000002010 (MNVsmall) is an engineering/ manoeuvre configuration; the same log marks its source outside the SXS field. It is retained as a separately labelled released product, not as a representative target observation.

from datasets import load_dataset
dataset = load_dataset("astro-legacy-archive/hitomi-sxs-cleaned-events", "ah100040010sxs_p0px1010_cl__EVENTS", split="train", streaming=True)
print(next(iter(dataset))["TIME"])

23196

import json
from huggingface_hub import hf_hub_download
import pyarrow.parquet as pq
path = hf_hub_download("astro-legacy-archive/hitomi-sxs-cleaned-events", "ah100040010sxs_p0px1010_cl__EVENTS.parquet", repo_type="dataset")
print(json.loads(pq.read_metadata(path).metadata[b"event_list"])["schema"])

astro-legacy-archive/event-list

Sources and structure

Canonical sources, checked 2026-08-30:

• —https://heasarc.gsfc.nasa.gov/FTP/hitomi/data/obs/1/100040010/sxs/event_cl/ah100040010sxs_p0px1010_cl.evt.gz — 1793063 bytes; SHA-256 57ad7ed1375326c931c4ff53fdac1ddc1cc854570f7f94c474490a0e4687b845
• —https://heasarc.gsfc.nasa.gov/FTP/hitomi/data/obs/0/000002010/sxs/event_cl/ah000002010sxs_p0px1010_cl.evt.gz — 14599 bytes; SHA-256 c8cb7adc3feef971cc6da04c442272a5e86a119a7ee666aad7f1fe5ec5001760

Both configurations mirror the released EVENTS table and its GTI companion verbatim. Both EVENTS tables contain the same 55 source columns in source order:

TIME 1D s; TRIGTIME 1D s; S_TIME 1D s; L32TI 1J; CATEGORY 1B; ADU_CNT 1B; PSP_ID 1B; FORMAT_VER 1B; WFRB_WRITE_LP 1J; WFRB_SAMPLE_CNT 1J; NUM_ELEM 1B; SUM_LOST_CNT 1J; ITYPE 1B; TYPE 2A; IPIX 1B; PIXEL 1B; TRIG_LP 1J; QUICK_DOUBLE 1X; SLOPE_DIFFER 1X; LO_RES_PH 1I; DERIV_MAX 1I; RISE_TIME 1B; TICK_SHIFT 1B; TIME_VERNIER 1B; PHA 1J; FLAGS 1B; EL_LOST_CNT 1I; EL_REASON 1B; EL_STOP_LP 1J; PREV_INTERVAL 1I; NEXT_INTERVAL 1I; SAMPLECNT 1D; SAMPLECNTTRIG 1D; ACTX 1B; ACTY 1B; DETX 1B; DETY 1B; FOCX 1I; FOCY 1I; X 1I; Y 1I; UPI 1E; EPI 1E; EPI2 1E; EPIPER 1E; EPI2PER 1E; PI 1J; INDEX 1J; GROUPS 1J; CTMULT 1I; STATUS 16X; PROC_STATUS 32X; CTMULT2 I; SEQ J; PHA2 J.

Each companion GTI table contains source-named START 1D s and STOP 1D s columns. Every Parquet file embeds an event-list 1.2.0 JSON document under the event_list metadata key. For a machine-written resolver, source column order, forms, units, scaling, null declarations, served encodings and bit mappings are at fields; the original per-column meanings are the TTYPE comments in the EVENTS header at fits_sources[0]["fits_headers"][1]["header"]; every GTI value is at auxiliary_tables; and all source headers are at fits_sources. The 16X and 32X values are fixed-size Boolean lists whose list element i is FITS nX element i in stored order. Byte, string, bit, scaled and unsigned physical values are not recast into new source identifiers.

event-list 1.1.1 did not fit because EVENTS uses B, A, 1X, 16X and explicit-repeat numeric forms. event-list 1.2.0 preserves those forms and the source-named GTI table verbatim.

The Perseus file contains 23,196 events and 19 GTIs: 22,742 Hp, 39 Mp, 33 Ms, 101 Lp and 281 Ls events. Its accepted exposure is 48,723.994761 s within an 86,007.556728 s outer span. The engineering file contains 37 events (35 Hp and 2 Lp); its two GTIs provide 1,108.5126333 s exposure across a 6,052.740733 s outer span. In that file PI reaches 32,768 despite the header-defined 0--32,767 range, and EPI ranges from 318.08282470703125 to 58470.09375 eV, beyond the nominal 0.3--12 keV SXS response. The Perseus file has PI from 602 to 32768 and EPI from 301.11895751953125 to 14246512 eV. These source values are preserved without clipping or reinterpretation; they are not evidence for useful calibrated response at those energies.

Interpretation and caveats

These tables cover the whole active 35-pixel science array/full SXS field, not photons extracted from the named source. The inactive corner science pixel 12 is replaced physically by the offset calibration pixel; it is not a 36th active imaging pixel. “Cleaned” is not “analysis-ready”: scientific use requires source and background regions; STATUS/PROC_STATUS, crosstalk, rise-time, PI, grade and pixel screening; grade-appropriate responses; and non-X-ray-background treatment. Perseus analysis also requires the mission's dedicated sxsperseus calibration workflow. Consult the official observation log and analysis guide and archive description.

Times are TT with TIMEREF=LOCAL and CLOCKAPP=T: spacecraft clock correction was applied, but the times are not barycentric. The embedded historical snapshot is PROCVER=03.01.006.007 and CALDBVER=gen20170208_hxi20161122_sgd20170208_sxi20170405_sxs20161122. HEASARC's later Hitomi CALDB v10 uses SXS CALDB 20180212 (released 2018-02-15), so the embedded snapshot is not current. Reproducible science should follow current Hitomi calibration guidance and begin from the unfiltered products where reprocessing is required. A census on 2026-08-30 found 42 archive observation directories; that is a dated archive count, not a mission-total claim.

Citation, licence and credit

Cite Takahashi, T. et al. 2018, Journal of Astronomical Telescopes, Instruments, and Systems 4, 021402, DOI 10.1117/1.JATIS.4.2.021402, and Kelley, R. L. et al. 2016, Proc. SPIE 9905, 99050V, DOI 10.1117/12.2232509.

No explicit data licence was found. HEASARC states:

HEASARC materials are all available freely for your use.

This is a free-use statement, not a named or formal licence. 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.

Errata and requests

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