Datasetpublic

AiCells/cells-demo-puller-csv-public

Cells demo puller CSV (synthetic, public) This is synthetic demo data for Cells pullers. Every brand, model name, price, store and date in it was invented by a seeded random generator (make_data.py, included). None of it describes a real product, company, price or person, and nothing was scraped or copied from anywhere. It exists so the Cells team can show and test a puller (Cells' coming data-import feature) that downloads a CSV from a Hugging Face dataset at a pinned revision… See the full description on the dataset page: https://huggingface.co/datasets/AiCells/cells-demo-puller-csv-public.

sourceHugging Facecc0-1.0updated 12d agoView on Hugging Face

0likes81downloads

Dataset Card

Cells demo puller CSV (synthetic, public)

This is synthetic demo data for Cells pullers. Every brand, model name, price, store and date in it was invented by a seeded random generator (make_data.py, included). None of it describes a real product, company, price or person, and nothing was scraped or copied from anywhere.

It exists so the Cells team can show and test a puller (Cells' coming data-import feature) that downloads a CSV from a Hugging Face dataset at a pinned revision, using the official huggingface_hub client, and emits the rows through the Cells puller SDK (cellib.puller_api).

Public mirror

This is the public copy of AiCells/cells-demo-puller-csv (identical files), so the Cells example puller works without a Hugging Face sign-in. The private original is the "with sign-in" variant of the same example.

Files

File	Rows	Columns
`hardware_prices.csv`	200	`id` (int), `brand`, `model`, `category` (cpu, gpu, ssd, ram, motherboard, psu, monitor), `price_usd` (2 decimals), `release_year` (2019–2026)
`price_observations.csv`	120	`observation_id`, `product_id` (→ `hardware_prices.id`), `store`, `observed_on` (ISO date), `price_usd` (empty when out of stock), `in_stock` (`true`/`false`)

The second file has a different shape on purpose (long format, dates, an empty cell and booleans), so a puller's schema handling can be tested on both.

Reproduce

bash

python make_data.py   # writes both CSVs, byte for byte

License

The data is synthetic and dedicated to the public domain under CC0 1.0 Universal (cc0-1.0). make_data.py is released under the same terms.

Access

The repository is private to the AiCells organization. Its tag v1.0 marks the revision the Cells demo pins.