Hugging Face's logo

Datasets:
policyengine
/
microcosm-us-sources

Tags:
microsimulation
census
cps-asec
License:
Dataset card Files
xet
 Community
Dataset Viewer
Duplicate
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code:   StreamingRowsError
Exception:    ArrowInvalid
Message:      Mismatching child array lengths
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 147, in get_rows_or_raise
                  return get_rows(
                      dataset=dataset,
                  ...<4 lines>...
                      column_names=column_names,
                  )
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                File "/src/services/worker/src/worker/utils.py", line 127, in get_rows
                  rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
                File "/src/services/worker/src/worker/utils.py", line 483, in safe_iter
                  yield from ds.decode(False) if ds.features else ds
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2840, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2398, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 536, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/hdf5/hdf5.py", line 83, in _generate_tables
                  pa_table = _recursive_load_arrays(h5, self.info.features, start, end)
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/hdf5/hdf5.py", line 267, in _recursive_load_arrays
                  arr = _recursive_load_arrays(obj, features[path], start, end)
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/hdf5/hdf5.py", line 288, in _recursive_load_arrays
                  sarr = pa.StructArray.from_arrays(values, names=keys)
                File "pyarrow/array.pxi", line 4304, in pyarrow.lib.StructArray.from_arrays
                File "pyarrow/error.pxi", line 155, in pyarrow.lib.pyarrow_internal_check_status
                File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
                  raise convert_status(status)
              pyarrow.lib.ArrowInvalid: Mismatching child array lengths

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.

Microcosm US build sources

Processed Current Population Survey Annual Social and Economic Supplement (CPS ASEC) files that the PolicyEngine Microcosm US build reads as inputs to its base stage (tools/build_us_puf_support_base.py --asec-h5 YEAR=PATH in PolicyEngine/microcosm).

The 2022-2024 files are mirrored here unchanged so that the build's inputs do not exist on one machine only. The 2025 file was built for this repository on 27 September 2026 (see Provenance). Microcosm pins each file by repository revision, SHA-256 and size, and refuses a file whose digest differs. Files are named by income year; the survey year is the income year plus one.

Files

File Survey year (income year) Bytes SHA-256 Households Persons
census_cps_2025.h5 ASEC 2026 (2025) 304,753,967 4c5a32188b6acfbcfbeb9f5d719d3847873887b72ebdbb19bc402c16e0a64e58 53,344 134,729
census_cps_2024.h5 ASEC 2025 (2024) 323,994,739 ec36604cb735a660b51b0b2f90be27d803b5878f3464fb30d0eacead59c1260d 55,762 142,125
census_cps_2023.h5 ASEC 2024 (2023) 299,036,610 cb57817327799f42b741caed5f9be94d04021c2e6809c1ad7bd0686da5428d88 56,251 144,265
census_cps_2022.h5 ASEC 2023 (2022) 301,129,278 7ccca976284bb47815d84460cc4f75a0a65d26d7754ab0a0f417de351b3d474e 56,839 146,133

For 2022-2024, household and person counts are the ones the Microcosm base stage recorded when it read these exact bytes on 16 September 2026. For 2025 they are the row counts of the file's household and person tables.

Provenance

Each file is an HDF5 store (pandas HDFStore, fixed format) of the person, family, household, SPM unit and tax unit tables of the Census Bureau's public-use CPS ASEC release for that year:

  • 2025: https://www2.census.gov/programs-surveys/cps/datasets/2026/march/asecpub26csv.zip (139,103,894 bytes, SHA-256 fe819d0d2fc4470c282e76c247b8aa8b6054811619730307e9f2fe6186707d93, Last-Modified 15 September 2026)
  • 2024: https://www2.census.gov/programs-surveys/cps/datasets/2025/march/asecpub25csv.zip
  • 2023: https://www2.census.gov/programs-surveys/cps/datasets/2024/march/asecpub24csv.zip
  • 2022: https://www2.census.gov/programs-surveys/cps/datasets/2023/march/asecpub23csv.zip

The 2022-2024 files were produced by PolicyEngine's retired US data package's Census CPS loader. The exact commit that wrote each file was not recorded, so these files are pinned by their bytes. The PolicyEngine organisation publishes earlier files with the same names in another repository; those have different bytes and are not interchangeable with these.

census_cps_2025.h5

Built by the same loader at PolicyEngine/policyengine-us-data@40314a75 ("Construct CPS tax units from household records", #824), plus one commit, MaxGhenis/policyengine-us-data@9f177ddd, that adds the income-year 2025 URL and stops requesting SPM_BBSUBVAL, which the ASEC 2026 person file no longer carries. The upstream repository is archived, so that commit lives on a fork.

That loader revision was chosen because, rebuilt from asecpub25csv.zip on 27 September 2026, it reproduces every table of census_cps_2024.h5 above exactly (same keys, columns, column order, dtypes, index and values). The rebuilt file's SHA-256 differs from the pinned one because HDF5 metadata is not byte-deterministic, which is why files here are pinned by the uploaded bytes rather than by a rebuild hash. The later archived loader head (42ed5d45) adds seven person columns the 2024 file lacks, so it was not used.

Relative to census_cps_2024.h5, the 2025 file's schema differs only where the Census files do:

  • person (161 columns) and spm_unit (38 columns) lack SPM_BBSUBVAL. It is omitted, not filled with zeros, because Census did not publish it for 2025.
  • household (153 columns) lacks HBBSUB_MNTH, HBBSUB_YN, HECC_PROBTIME, HXCC_PROBTIME, I_HBBSUBMNTH and I_HBBSUBYN, and adds HTCC_PROBLOSS and HXCC_PROBLOSS, as hhpub26.csv does.
  • family (85 columns) and tax_unit (1 column) have the same columns.

All 18 NOW_* at-interview coverage recodes the 2024 file carries are present with codes {1, 2}.

Licence and source terms

The underlying data are public-use microdata from the US Census Bureau and are in the public domain in the United States. PolicyEngine's processed files are offered under CC BY 4.0. The Census Bureau does not endorse this product.
Downloads last month
70
Total file size:
1.23 GB