Datasets:

SabaPivot
/

icml26-repro-mean-shift-pca

Modalities:

Dataset Preview

Split (1)

train

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 2 new columns ({'edge', 'spike'}) and 4 missing columns ({'stable_count', 'method', 'pi', 'alignment'}).

This happened while the csv dataset builder was generating data using

hf://datasets/SabaPivot/icml26-repro-mean-shift-pca/results/fluctuations.csv (at revision c2a2103e95ef98b2cb144d0336bd8b98dd0ac7fa), ['hf://datasets/SabaPivot/icml26-repro-mean-shift-pca@c2a2103e95ef98b2cb144d0336bd8b98dd0ac7fa/results/algorithm_benchmark.csv', 'hf://datasets/SabaPivot/icml26-repro-mean-shift-pca@c2a2103e95ef98b2cb144d0336bd8b98dd0ac7fa/results/fluctuations.csv', 'hf://datasets/SabaPivot/icml26-repro-mean-shift-pca@c2a2103e95ef98b2cb144d0336bd8b98dd0ac7fa/results/invariance.csv', 'hf://datasets/SabaPivot/icml26-repro-mean-shift-pca@c2a2103e95ef98b2cb144d0336bd8b98dd0ac7fa/results/spectral_separation.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 1837, 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 2369, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2297, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              n: int64
              rep: int64
              edge: double
              spike: double
              -- schema metadata --
              pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 700
              to
              {'n': Value('int64'), 'pi': Value('float64'), 'rep': Value('int64'), 'method': Value('string'), 'alignment': Value('float64'), 'stable_count': Value('float64')}
              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 1683, 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 1839, 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 2 new columns ({'edge', 'spike'}) and 4 missing columns ({'stable_count', 'method', 'pi', 'alignment'}).
              
              This happened while the csv dataset builder was generating data using
              
              hf://datasets/SabaPivot/icml26-repro-mean-shift-pca/results/fluctuations.csv (at revision c2a2103e95ef98b2cb144d0336bd8b98dd0ac7fa), ['hf://datasets/SabaPivot/icml26-repro-mean-shift-pca@c2a2103e95ef98b2cb144d0336bd8b98dd0ac7fa/results/algorithm_benchmark.csv', 'hf://datasets/SabaPivot/icml26-repro-mean-shift-pca@c2a2103e95ef98b2cb144d0336bd8b98dd0ac7fa/results/fluctuations.csv', 'hf://datasets/SabaPivot/icml26-repro-mean-shift-pca@c2a2103e95ef98b2cb144d0336bd8b98dd0ac7fa/results/invariance.csv', 'hf://datasets/SabaPivot/icml26-repro-mean-shift-pca@c2a2103e95ef98b2cb144d0336bd8b98dd0ac7fa/results/spectral_separation.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.

n int64	pi float64	rep int64	method string	alignment float64	stable_count float64
160	0.05	0	MS-PCA	0.979876	3
160	0.05	0	PCA	0.170585	null
160	0.05	0	centered PCA	0.124758	null
160	0.05	1	MS-PCA	0.987034	4
160	0.05	1	PCA	0.043731	null
160	0.05	1	centered PCA	0.052883	null
160	0.05	2	MS-PCA	0.815456	3
160	0.05	2	PCA	0.57296	null
160	0.05	2	centered PCA	0.588604	null
160	0.05	3	MS-PCA	0.993828	3
160	0.05	3	PCA	0.069727	null
160	0.05	3	centered PCA	0.084264	null
160	0.05	4	MS-PCA	0.847775	3
160	0.05	4	PCA	0.493068	null
160	0.05	4	centered PCA	0.508732	null
160	0.05	5	MS-PCA	0.985189	3
160	0.05	5	PCA	0.121835	null
160	0.05	5	centered PCA	0.13752	null
160	0.05	6	MS-PCA	0.997912	3
160	0.05	6	PCA	0.00021	null
160	0.05	6	centered PCA	0.027874	null
160	0.05	7	MS-PCA	0.957504	3
160	0.05	7	PCA	0.071127	null
160	0.05	7	centered PCA	0.089254	null
160	0.05	8	MS-PCA	0.95388	3
160	0.05	8	PCA	0.182204	null
160	0.05	8	centered PCA	0.194758	null
160	0.05	9	MS-PCA	0.991656	4
160	0.05	9	PCA	0.008411	null
160	0.05	9	centered PCA	0.002432	null
160	0.05	10	MS-PCA	0.008699	2
160	0.05	10	PCA	0.194372	null
160	0.05	10	centered PCA	0.29294	null
160	0.05	11	MS-PCA	0.844591	3
160	0.05	11	PCA	0.346631	null
160	0.05	11	centered PCA	0.352357	null
160	0.25	0	MS-PCA	0.97569	4
160	0.25	0	PCA	0.187525	null
160	0.25	0	centered PCA	0.056989	null
160	0.25	1	MS-PCA	0.985005	3
160	0.25	1	PCA	0.059174	null
160	0.25	1	centered PCA	0.102822	null
160	0.25	2	MS-PCA	0.04209	2
160	0.25	2	PCA	0.60099	null
160	0.25	2	centered PCA	0.729277	null
160	0.25	3	MS-PCA	0.018876	1
160	0.25	3	PCA	0.349089	null
160	0.25	3	centered PCA	0.504572	null
160	0.25	4	MS-PCA	0.958531	2
160	0.25	4	PCA	0.228122	null
160	0.25	4	centered PCA	0.355955	null
160	0.25	5	MS-PCA	0.994552	3
160	0.25	5	PCA	0.051209	null
160	0.25	5	centered PCA	0.213313	null
160	0.25	6	MS-PCA	0.993481	3
160	0.25	6	PCA	0.077639	null
160	0.25	6	centered PCA	0.232543	null
160	0.25	7	MS-PCA	0.981089	3
160	0.25	7	PCA	0.041993	null
160	0.25	7	centered PCA	0.128304	null
160	0.25	8	MS-PCA	0.936096	3
160	0.25	8	PCA	0.242019	null
160	0.25	8	centered PCA	0.383563	null
160	0.25	9	MS-PCA	0.988602	4
160	0.25	9	PCA	0.085207	null
160	0.25	9	centered PCA	0.156933	null
160	0.25	10	MS-PCA	0.997729	2
160	0.25	10	PCA	0.042956	null
160	0.25	10	centered PCA	0.132416	null
160	0.25	11	MS-PCA	0.093749	2
160	0.25	11	PCA	0.229732	null
160	0.25	11	centered PCA	0.286114	null
260	0.05	0	MS-PCA	0.853241	2
260	0.05	0	PCA	0.330167	null
260	0.05	0	centered PCA	0.337246	null
260	0.05	1	MS-PCA	0.011838	1
260	0.05	1	PCA	0.353046	null
260	0.05	1	centered PCA	0.32943	null
260	0.05	2	MS-PCA	0.983353	3
260	0.05	2	PCA	0.105654	null
260	0.05	2	centered PCA	0.074269	null
260	0.05	3	MS-PCA	0.016921	3
260	0.05	3	PCA	0.030067	null
260	0.05	3	centered PCA	0.011177	null
260	0.05	4	MS-PCA	0.952732	3
260	0.05	4	PCA	0.000883	null
260	0.05	4	centered PCA	0.041909	null
260	0.05	5	MS-PCA	0.963594	3
260	0.05	5	PCA	0.20575	null
260	0.05	5	centered PCA	0.243549	null
260	0.05	6	MS-PCA	0.968797	4
260	0.05	6	PCA	0.099844	null
260	0.05	6	centered PCA	0.120234	null
260	0.05	7	MS-PCA	0.988312	3
260	0.05	7	PCA	0.094286	null
260	0.05	7	centered PCA	0.100133	null
260	0.05	8	MS-PCA	0.181735	2
260	0.05	8	PCA	0.060801	null
260	0.05	8	centered PCA	0.057208	null
260	0.05	9	MS-PCA	0.70913	2

End of preview.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

Mean-Shift PCA reproduction

Independent high-dimensional Gaussian-mixture checks for spectral separation, eigenspace residuals, fluctuation orders, and Algorithm 1 at 5% contamination. Run uv run --with-requirements requirements.txt python reproduce.py.

Downloads last month: 7

Total file size:

1.17 MB