The dataset viewer is not available for this split.
Error code: UnexpectedError
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
japan_ndb_open — mirror
A convenience mirror of a public dataset, pushed by propensity.data.hf_mirror so that benchmark runs do not depend on a flaky upstream origin.
This is a mirror, not a grant of rights
Copyright and licence remain with the original source named below. Mirroring moves bytes; it does not and cannot change the terms they arrived under. If you intend to redistribute this further, or to use it commercially, read the original terms — not this card, and not the license: field above, which is our best reading and may be wrong.
Provenance
- Source: Source: 12th NDB Open Data, medical procedures by secondary medical area, Ministry of Health, Labour and Welfare (Japan), adapted. Licensed under the Public Data License 1.0 (PDL1.0), which the Digital Agency states is CC BY 4.0 compatible and which expressly permits commercial use. MHLW's grant is CONDITIONAL on there being no in-file rights notice; both that condition (all 60 workbooks, 180 package parts, scanned -- zero notices) and the enumerated carve-out list (five items, none of them this dataset) were TESTED, not assumed. FOUR CAVEATS. (1) THERE IS NO MONETARY COLUMN. The tariff is in NHI POINTS, not yen; points convert to yen only through a statutory multiplier that is not in the file and is not applied here. Do not present tariff points as money or sum them as revenue. (2) The entity axis is PROCEDURE x GEOGRAPHY only -- no payer, insurer, provider, plan or person, and no adjudication, denial, enrolment or premium. This is a FEATURE source (a national price-and-volume covariate) and must never be an insurance outcome. (3) The time index is parsed out of each sheet's A1 title string; there is no date column anywhere in the file. (4) THE NATIONAL TOTAL IS A LEVEL OF THE AREA DIMENSION, flagged -- an unfiltered sum is 2.002x the truth, measured. And the geography has 335 real secondary medical areas plus one unassignable residual, not 336 places. SUPPRESSION IS SUBTLER THAN DOCUMENTED: the published marker is U+002D HYPHEN-MINUS while the file's own note uses U+2010 HYPHEN, so a loader trusting the documentation finds zero suppression; and the rule is COMPLEMENTARY as well as primary -- if one cell in a row falls below the threshold, every cell holding the smallest value at or above it is masked too, so a withheld cell is NOT necessarily small. The threshold is also not uniform across chapters. WHAT IT IS NOT: not claim-level, not monetary, not an outcome source, and not a grant of rights by this repository.
- Origin URL: https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000177221_00020.html
- Loader:
propensity.data.public_loaders.load_japan_ndb_open - Extract: full pull from origin
- Reported source:
japan_ndb_open - Rows pushed: 987,073
- Columns: 26
Licence and redistribution
- Licence as found:
other(jp-pdl-1.0) - Redistribution verdict:
open - Basis: Japan MHLW, Public Data License 1.0 (PDL1.0), CC BY 4.0 compatible. MHLW's terms page https://www.mhlw.go.jp/chosakuken/index.html (HTTP 200, read 2026-11-10) grants PDL1.0 to its content CONDITIONAL on there being no in-file rights notice. PDL1.0's own text, from its maintainer the Digital Agency, permits reproduction, public transmission and adaptation and states commercial use is possible, and its clause 1.7 states the licence is CC BY 4.0 compatible and that the State permits users to use the content under CC BY. BOTH CONDITIONS WERE TESTED, which is the point. (a) The in-file rights notice: xl/sharedStrings.xml, docProps/core.xml and docProps/app.xml of ALL 60 workbooks (180 parts) were scanned for rights tokens -- zero notices, dc:creator empty. That widens the sweep's single-workbook check to the whole archive. (b) The carve-out list the sweep did NOT test: MHLW says content under a different rule is enumerated on a separate page, and that page (HTTP 200, and served as Shift_JIS while the terms page is UTF-8 -- a decode trap) lists exactly five items: the MHLW symbol mark, the Kirara mark, the Tomonin mark, the Labour Relations Commission case database, and Central Labour Relations Commission notices. NDB Open Data is not among them. What travels: the source statement in MHLW's own template, a statement that the content was edited/processed and by whom, no implication that the Japanese state produced the derivative, and the third-party-rights caveat. No personal data: procedure x region billing counts.
Citation
出典:「第12回NDBオープンデータ 医科診療行為(算定回数)二次医療圏別」(厚生労働省)(https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000177221_00020.html)(2026年11月10日に利用)を加工して作成。PDL1.0 (https://www.digital.go.jp/resources/open_data/public_data_license_v1.0). English: Ministry of Health, Labour and Welfare (2025), 12th NDB Open Data, medical procedure billed counts by secondary medical area; reshaped from the published workbooks by this loader. Licensed under PDL1.0, which the Digital Agency states is compatible with CC BY 4.0.
Loading
from propensity.data.hf_mirror import load_mirrored
result = load_mirrored("japan_ndb_open", org="SMVaddadi")
result.frame.head()
load_mirrored falls back to the original loader (origin, then its own offline replica) if this mirror is unreachable, so it is safe to depend on.
- Downloads last month
- 52