Datasets:

Omcrec
/

ecommerce-ai-data-analyst-agent-benchmark

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 4 new columns ({'signup_date', 'customer_id', 'acquisition_channel', 'region'}) and 3 missing columns ({'date', 'spend', 'channel'}).

This happened while the csv dataset builder was generating data using

hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark/customers.csv (at revision 4536d2e001f407cf37d4b607070dedeb0bac4ddd), ['hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark@4536d2e001f407cf37d4b607070dedeb0bac4ddd/ad_spend.csv', 'hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark@4536d2e001f407cf37d4b607070dedeb0bac4ddd/customers.csv', 'hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark@4536d2e001f407cf37d4b607070dedeb0bac4ddd/inventory.csv', 'hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark@4536d2e001f407cf37d4b607070dedeb0bac4ddd/orders.csv', 'hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark@4536d2e001f407cf37d4b607070dedeb0bac4ddd/products.csv', 'hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark@4536d2e001f407cf37d4b607070dedeb0bac4ddd/returns.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
              customer_id: string
              signup_date: string
              region: string
              acquisition_channel: string
              -- schema metadata --
              pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 772
              to
              {'date': Value('string'), 'channel': Value('string'), 'spend': 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 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 4 new columns ({'signup_date', 'customer_id', 'acquisition_channel', 'region'}) and 3 missing columns ({'date', 'spend', 'channel'}).
              
              This happened while the csv dataset builder was generating data using
              
              hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark/customers.csv (at revision 4536d2e001f407cf37d4b607070dedeb0bac4ddd), ['hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark@4536d2e001f407cf37d4b607070dedeb0bac4ddd/ad_spend.csv', 'hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark@4536d2e001f407cf37d4b607070dedeb0bac4ddd/customers.csv', 'hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark@4536d2e001f407cf37d4b607070dedeb0bac4ddd/inventory.csv', 'hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark@4536d2e001f407cf37d4b607070dedeb0bac4ddd/orders.csv', 'hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark@4536d2e001f407cf37d4b607070dedeb0bac4ddd/products.csv', 'hf://datasets/Omcrec/ecommerce-ai-data-analyst-agent-benchmark@4536d2e001f407cf37d4b607070dedeb0bac4ddd/returns.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.

date string	channel string	spend float64
2026-01-01	Paid Search	615.07
2026-01-01	Social	314.36
2026-01-01	Display	267.71
2026-01-01	Affiliate	186.98
2026-01-02	Paid Search	670.06
2026-01-02	Social	381.39
2026-01-02	Display	213.85
2026-01-02	Affiliate	191.72
2026-01-03	Paid Search	510.66
2026-01-03	Social	339.72
2026-01-03	Display	260.69
2026-01-03	Affiliate	164.09
2026-01-04	Paid Search	457.81
2026-01-04	Social	419.06
2026-01-04	Display	292.14
2026-01-04	Affiliate	188.37
2026-01-05	Paid Search	506.25
2026-01-05	Social	442.63
2026-01-05	Display	223.03
2026-01-05	Affiliate	181.11
2026-01-06	Paid Search	669
2026-01-06	Social	366.65
2026-01-06	Display	283.37
2026-01-06	Affiliate	174.36
2026-01-07	Paid Search	636.52
2026-01-07	Social	309.51
2026-01-07	Display	301.38
2026-01-07	Affiliate	173.83
2026-01-08	Paid Search	585.91
2026-01-08	Social	424.02
2026-01-08	Display	276.55
2026-01-08	Affiliate	178.35
2026-01-09	Paid Search	604.04
2026-01-09	Social	352.96
2026-01-09	Display	272.41
2026-01-09	Affiliate	219.84
2026-01-10	Paid Search	486.26
2026-01-10	Social	416.54
2026-01-10	Display	306.76
2026-01-10	Affiliate	145.69
2026-01-11	Paid Search	656.17
2026-01-11	Social	324.54
2026-01-11	Display	267.95
2026-01-11	Affiliate	148.83
2026-01-12	Paid Search	632.72
2026-01-12	Social	364.79
2026-01-12	Display	277.27
2026-01-12	Affiliate	175.62
2026-01-13	Paid Search	487.63
2026-01-13	Social	459.75
2026-01-13	Display	257.6
2026-01-13	Affiliate	204.31
2026-01-14	Paid Search	642.68
2026-01-14	Social	452.32
2026-01-14	Display	207.07
2026-01-14	Affiliate	211.11
2026-01-15	Paid Search	476.16
2026-01-15	Social	487.45
2026-01-15	Display	273.55
2026-01-15	Affiliate	163.49
2026-01-16	Paid Search	485.37
2026-01-16	Social	435.54
2026-01-16	Display	203.38
2026-01-16	Affiliate	154
2026-01-17	Paid Search	533.29
2026-01-17	Social	447.39
2026-01-17	Display	242.56
2026-01-17	Affiliate	166.41
2026-01-18	Paid Search	446.58
2026-01-18	Social	438.99
2026-01-18	Display	214.62
2026-01-18	Affiliate	169.19
2026-01-19	Paid Search	583.72
2026-01-19	Social	451.26
2026-01-19	Display	198.23
2026-01-19	Affiliate	137.9
2026-01-20	Paid Search	562.08
2026-01-20	Social	445.86
2026-01-20	Display	297.13
2026-01-20	Affiliate	174.66
2026-01-21	Paid Search	657.39
2026-01-21	Social	441.96
2026-01-21	Display	239.62
2026-01-21	Affiliate	195.06
2026-01-22	Paid Search	630.02
2026-01-22	Social	395.97
2026-01-22	Display	227.07
2026-01-22	Affiliate	199.11
2026-01-23	Paid Search	605.15
2026-01-23	Social	402.68
2026-01-23	Display	253.24
2026-01-23	Affiliate	221.36
2026-01-24	Paid Search	556.22
2026-01-24	Social	403
2026-01-24	Display	307.67
2026-01-24	Affiliate	195.64
2026-01-25	Paid Search	589.79
2026-01-25	Social	476.84
2026-01-25	Display	232.66
2026-01-25	Affiliate	167.61

End of preview.

E-commerce AI Data Analyst Agent Benchmark

A synthetic e-commerce dataset for evaluating AI data analyst agents on realistic, multi-step business analysis, data-quality investigation, and analytical reasoning.

This dataset is part of the E-commerce AI Data Analyst Agent Benchmark.

Dataset summary

This dataset supports evaluation of AI data analyst agents on realistic, multi-step e-commerce analysis.

It contains:

customers.csv
products.csv
orders.csv
returns.csv
ad_spend.csv
inventory.csv

The benchmark covers tasks involving:

data retrieval and aggregation
business analysis
data-quality investigation
reconciliation
analytical reasoning
business interpretation

Data provenance

The datasets are synthetic and were generated specifically for this benchmark. They are not sourced from a real company's customer or transaction database.

The data intentionally contains controlled defects for some benchmark tasks, including duplicate records, missing or orphan references, invalid numeric values, and return-linkage problems.

Intended use

This dataset is intended for:

AI-agent evaluation
data analyst agent benchmarking
research and experimentation
reproducible testing of analytical workflows
educational and noncommercial projects

Limitations

This is a synthetic benchmark dataset. Results obtained using it should not be interpreted as evidence of performance on any particular real-world company or production dataset.

The dataset is designed for benchmark evaluation and contains deliberately constructed data-quality defects. Those defects should not be treated as representative estimates of real-world data quality.

Related benchmark

The complete benchmark, including tasks, ground truth, evaluator, adversarial cases, calibration tests, pilot results, and documentation is available in the GitHub repository:

https://github.com/Omcrec/ecommerce-ai-data-analyst-agent-benchmark

License

The dataset and associated benchmark content in this repository are licensed under the Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0).

You may copy, modify, remix, and redistribute the material for noncommercial purposes with appropriate attribution.

Commercial use, commercial distribution, or sale of the dataset or substantial portions of its content is not permitted under this license.

License: https://creativecommons.org/licenses/by-nc/4.0/

Version

Benchmark version: 1.0.0

Downloads last month: 98

Total file size:

122 kB