Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed because of a cast error
Error code: DatasetGenerationCastError
Exception: DatasetGenerationCastError
Message: An error occurred while generating the dataset
All the data files must have the same columns, but at some point there are 9 new columns ({'n_slices', 'label', 'n_pos_slices', 'topfrac_mean', 'study', 'any_slice_flag', 'max', 'mean', 'topk_mean'}) and 3 missing columns ({'hash', 'any_positive', 'ID'}).
This happened while the csv dataset builder was generating data using
hf://datasets/Obaidullahmiakhil/ich_audit-ef03/tables/expG_studies_test.csv (at revision 799e9b1d45ddb3e0b12ac6c4009692f55025e0e2), ['hf://datasets/Obaidullahmiakhil/ich_audit-ef03@799e9b1d45ddb3e0b12ac6c4009692f55025e0e2/tables/exp0_leaky_test_ids.csv', 'hf://datasets/Obaidullahmiakhil/ich_audit-ef03@799e9b1d45ddb3e0b12ac6c4009692f55025e0e2/tables/expG_studies_test.csv']
Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1848, in _prepare_split_single
writer.write_table(table)
~~~~~~~~~~~~~~~~~~^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 765, in write_table
self._write_table(pa_table, writer_batch_size=writer_batch_size)
~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 773, in _write_table
pa_table = table_cast(pa_table, self._schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
study: string
n_slices: int64
label: int64
n_pos_slices: int64
max: double
topk_mean: double
topfrac_mean: double
mean: double
any_slice_flag: double
-- schema metadata --
pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 1320
to
{'ID': Value('string'), 'hash': Value('string'), 'any_positive': Value('int64')}
because column names don't match
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1369, in compute_config_parquet_and_info_response
parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
~~~~~~~~~~~~~~~~~~~~~~~~~^
builder, max_dataset_size_bytes=max_dataset_size_bytes
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 948, in stream_convert_to_parquet
builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1694, in _prepare_split
for job_id, done, content in self._prepare_split_single(
~~~~~~~~~~~~~~~~~~~~~~~~~~^
gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
):
^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1850, in _prepare_split_single
raise DatasetGenerationCastError.from_cast_error(
...<4 lines>...
)
datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
All the data files must have the same columns, but at some point there are 9 new columns ({'n_slices', 'label', 'n_pos_slices', 'topfrac_mean', 'study', 'any_slice_flag', 'max', 'mean', 'topk_mean'}) and 3 missing columns ({'hash', 'any_positive', 'ID'}).
This happened while the csv dataset builder was generating data using
hf://datasets/Obaidullahmiakhil/ich_audit-ef03/tables/expG_studies_test.csv (at revision 799e9b1d45ddb3e0b12ac6c4009692f55025e0e2), ['hf://datasets/Obaidullahmiakhil/ich_audit-ef03@799e9b1d45ddb3e0b12ac6c4009692f55025e0e2/tables/exp0_leaky_test_ids.csv', 'hf://datasets/Obaidullahmiakhil/ich_audit-ef03@799e9b1d45ddb3e0b12ac6c4009692f55025e0e2/tables/expG_studies_test.csv']
Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)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.
ID string | hash string | any_positive int64 |
|---|---|---|
ID_95e05fc39 | 5db0f05322d5bde9c814bd4c4fbb4ed1 | 1 |
ID_95f30ae5d | bea78338da84b4a909bdbccba8d4a8ed | 0 |
ID_d4b1e742e | e65abbfb256c3b5cb682606283d1f18e | 0 |
ID_e69ed1153 | 540c9af810cf69a177af4184308cb06b | 0 |
ID_ffb1b1df5 | 74f1f90432f78612e8d9fae44ec87eb4 | 1 |
ID_4ee92912b | 34131a3d9a6765ab9b5d1dec29cf28c5 | 0 |
ID_dc1bde73e | 4246c031463fb9b85db8f7ad3fd1d611 | 1 |
ID_21b1df599 | c96cf4e416b62e076f9070abb0a4325a | 1 |
ID_c1b1949cb | aea47aa3c5f29939d4dbcbfef735155f | 0 |
ID_5574dc93b | 4980eb595b37218793c31d01fdd8b09f | 1 |
ID_bf9ad4efe | 5905620775dc2c7d7ce9d42e258e80ef | 1 |
ID_64ff7b585 | d6f2e7ff8dbc2bc5eaab1711d718bad8 | 1 |
ID_c96df10ec | 115b7e221c244560418d8cd4c6e33a75 | 0 |
ID_fc31ad6f2 | d9abc9aff1fcb2653ddf39435c1d1c68 | 1 |
ID_f1d30159d | 56b222993c51558046a2ec000ab9fc51 | 1 |
ID_1fc866125 | 08bc23706359c356cc254549d4fde2c8 | 1 |
ID_871f11de8 | fabee87d168ee3c491d487be380f4616 | 0 |
ID_771a56452 | 094054a42d4fab6680c9f81b3a27ded9 | 0 |
ID_ab52c82c8 | c745e81032683a88beee71845d418913 | 0 |
ID_efd171694 | c99e5351283d951fa7f1dedd39ca08c6 | 0 |
ID_679748457 | 4c0efdceda4e5389a5b3a971a6d12f75 | 1 |
ID_f756f79b3 | a2a47ad4760c85ef3a7b48b32645422f | 1 |
ID_4b4b7bf63 | 314cf273f138c3ec0d298546a4c82bcc | 1 |
ID_30ad8e34a | d8478bb952a371675dd2b28e8be1079f | 0 |
ID_1b59c638b | ed82fca041ebffb7da55e7a415289dd9 | 0 |
ID_5b53fa247 | 55706e62eabcdf9581273af1ff57ca58 | 1 |
ID_ce8054486 | d94ae07deebf48dddbfaade13b708a5f | 1 |
ID_1150eb28d | 2cc3de89a0c270d6bdaaeadfdcbcd4b9 | 1 |
ID_53557cf18 | 3c5272fe890851eb5a787c8ff3c299c3 | 0 |
ID_04b6587ef | 6fa89d95942be6d17686eccf236d3a91 | 1 |
ID_29bab7a7f | 0631a1de11823adcc2322ed3722f765c | 1 |
ID_b917ae580 | 618ab2eb3b0ec259fdd598acedd430cf | 1 |
ID_aedb170f1 | 4823133a15f2a2c809b7187ef4eaec16 | 1 |
ID_c7fd531c1 | 120a5d862eb70e3eab8c61eb098e9b87 | 0 |
ID_5cfadca45 | 9b2abe0482fdc78ce5efad12c875f9db | 0 |
ID_ef6b43bb5 | 7f713580c413165264038a9a9ba93dc6 | 0 |
ID_7d45af173 | 2ecf8afd07f590309617ab5c243e52a1 | 1 |
ID_1eb191760 | ca3228a9b5895cfb44171817dd56883c | 0 |
ID_764776a9f | 69d0110665d7a3d9623ab7f29965d141 | 1 |
ID_2f2651c5e | 646e771970232acbe7095b54e8d7c8a9 | 0 |
ID_7aee3eac5 | eeb55ddd9572606661994d2e01a9ecf2 | 1 |
ID_aa8e4a60d | 5921d2731c2959f7dee5b0bc7d28c172 | 0 |
ID_37f5479e3 | 07afe956ec41f2dda1368c758efdf8f5 | 0 |
ID_6eadf4210 | d6286193e8901dbf3989aca3cd741239 | 0 |
ID_87a3d8c78 | ce3effdef3f249f3e6c1b11716574844 | 1 |
ID_41877cfdd | ea0add3fbd7c8c8d206c2a5961d9b688 | 0 |
ID_77da343b9 | 8ab2b3038ef89c31e39ec23dfc3f0ac2 | 0 |
ID_6ae182181 | 756fb805620dd22c49b3b3c1cf0c1d9e | 1 |
ID_9f354d90a | 80c05b5188abb3e9c4506afea3c51424 | 1 |
ID_ec82adc81 | 86575db1ceca059a69ad3862782b8c15 | 1 |
ID_64529ea8d | e9e06bb9f37a6f852960329a7c86dc9d | 1 |
ID_67309fa28 | 4230c2186c5e74d1f34dc89e4c6f0a0b | 1 |
ID_c9aa73bdd | 87ca81e65ac892f301d273dd5a7b480a | 1 |
ID_d1139ba99 | 956c9e3e6f4f635ccf42fc1136d3bc56 | 1 |
ID_999c98e60 | acce8bb3fe13151844a3dbfbb7b7daa3 | 0 |
ID_353960d3c | 472e892bc51dfddca0914adbd456dc87 | 0 |
ID_61e8d11f4 | 0eedd38358576045fa6cdce5b5b19a17 | 1 |
ID_12450592a | cd370b35a2e4a0d0e099f71de5c34336 | 1 |
ID_13c65999d | 24c454ee00da518ae0ae638559d6fc85 | 1 |
ID_8fb2230b4 | 16a8ab427839d2ed7a0c399608975334 | 0 |
ID_415de7e69 | 21f087adbdcdcab3538197d8bd4dc080 | 1 |
ID_4ebd2f3c2 | 684b7f292743c969231334f86af814af | 0 |
ID_7abbba029 | 4959a823388217b33386aa5ca3f9fa5b | 0 |
ID_0c6dd5075 | 96787252e416616ca76fe59a905c8483 | 1 |
ID_d11b20c3b | 018d6dffe0be7e8d8c08445d50f2a637 | 0 |
ID_6e291d211 | cbd885f3a0ed3503a3f7e7d4ac7c6b39 | 1 |
ID_aca6c319b | c1ff4fd22f4613d0001316b5e3f619a1 | 1 |
ID_a11571de2 | 49217f967d1d472b6c67202cc5353c0e | 0 |
ID_c2d72971c | bc3537705e1319b42b234390a51301d8 | 1 |
ID_fe5dc8ca3 | 2bfc286c2068a876d5751dddd8d39032 | 1 |
ID_becd61d14 | 100d01083c6ee4cb91910f0d9581684e | 0 |
ID_84fde1006 | beaa7e932b8cb910a4597e8ca4d5e545 | 0 |
ID_f3195854a | 5bd90cd30567f1133e9e29fbf27cd156 | 1 |
ID_62951fca5 | 4265ef2915a8f15253c92080038be24f | 0 |
ID_373297e76 | c8cb32c4adc0a1f6b5be1e175f916098 | 0 |
ID_001aba272 | 822cb219c7c3fe85243f0a9615933884 | 0 |
ID_683d9cb35 | 1979b20e8cb3565d7b722d17445a0d7b | 0 |
ID_917044c60 | a91865982f8ccd2ac8a826e10ec09752 | 0 |
ID_bbdd22ecb | e28694d140d4d6ee038e1f811e6fa008 | 0 |
ID_240e4ff95 | 10f43741592b7f8493c5a203c6d4710f | 0 |
ID_a71975eb2 | bbd40703e1913f72a89baa7faad7ba2c | 0 |
ID_d5c5c9eeb | fe2af324b62753b076091173bfc8cbdb | 1 |
ID_d8b410fa9 | 2897b6750a5925f46ba41824a7d27916 | 0 |
ID_0132ebd47 | 3a2bca81bdea2336e44aa91832026cca | 1 |
ID_f51d6077c | 35a6772ff71ef959a9eace411835c835 | 1 |
ID_80c5e2de0 | 56a5acd5877a631d199f4dba17525dd4 | 1 |
ID_6cee3062a | 53fea854e5d0d5fc29a48a99855162fa | 0 |
ID_63020644c | 2468d197ef47b8d410148283e3d3b03c | 1 |
ID_5e561dca3 | 70fe0551be77f0f66b828d90c30eb540 | 0 |
ID_105a89589 | c8e3d545ed28d0d6ab3425edca17fe62 | 0 |
ID_04dc78074 | 3609f8ebe274c270eeb1b62994a37fa2 | 0 |
ID_256c49e34 | edda5177ace1a0875806f67f01f1793b | 0 |
ID_5a66d02b3 | acdd25470aea61f0e975dc28ec41023c | 1 |
ID_7c09e66be | 57ecd79c94975d8393738000bdd288de | 0 |
ID_42cfdba3b | 247945f91a18b47c9dcdb0c80c03c846 | 0 |
ID_df3184cf2 | d30db7ac6c6cb8da4f400e8b735563ab | 1 |
ID_6022e7097 | e3fb4206070f6d15393ad58e9c930268 | 1 |
ID_2f5dc03cb | 88937f5bd1cc4f638327fa34dc60a94b | 0 |
ID_a87004f7a | ac4e950b72c83a3ddea5774b0d28ee0b | 0 |
ID_d1d9859c6 | 90a7aac1d4b92af3e1f9158dedc0a2b5 | 1 |
End of preview.
ICH EF03 — Audit bundle
Independent re-analysis of phase2-effnetb3-bigru-ef3-v1 (EfficientNet-B3 → BiGRU cascade with
conformal risk control, RSNA-2019 Intracranial Hemorrhage Detection).
No model was retrained. Everything here is a re-analysis of the existing Step-1 manifest and Step-2 per-slice score dumps.
| Experiment | Question | Verdict |
|---|---|---|
| 0 | Is the manifest + image store complete and consistent? | PASS |
| F | Does conformal risk control hold with a disjoint calibration set? | FAIL |
| G | Does the cascade triage studies, not just slices? | PASS |
Contents
EF03_audit.md full 22-section engineering + scientific audit
FINDINGS.md results of Experiments 0, F, G
results.json every number, machine-readable
MANIFEST.json file listing with sizes and sha256
split_repair_map.csv.gz ID -> (disk_chunk, disk_split): the Step-2 fix
labels_v4_fixed.parquet full manifest + disk_split/disk_chunk/path_ok
labels_v4_fixed.csv.gz same, gzipped CSV
figures/ PDF + PNG, 300 dpi
tables/ CSV for each analysis
Using the repair
split_repair_map.csv.gz fixes the path mismatch without touching a label or moving a
file. In Step 2, resolve images by disk_split rather than split:
rep = pd.read_csv("split_repair_map.csv.gz") # ID, chunk_id, split, disk_chunk, disk_split, path_ok
loc = dict(zip(rep.ID, zip(rep.disk_chunk, rep.disk_split)))
def resolve(image_id):
ch, sp = loc[image_id] # where the PNG ACTUALLY lives
return CHUNK_DIRS[ch] / sp / f"{image_id}.png"
The train/val/test assignment in split is unchanged and remains authoritative for
modelling; disk_split is a lookup detail only.
Reproducing
Run step4_audit_ef03.ipynb on Kaggle with the Step-1 manifest dataset, the Step-2
score dump, and (for Experiment 0) the ten image chunks and the raw RSNA DICOM dataset
attached.
Analysis version step4-audit-v1.
- Downloads last month
- 172