THULab/zillow_viewer

Housing Data Provided by Zillow (TsFile format)

This dataset contains seven products created from raw Zillow data published through the Zillow Research data program. The products cover market listings, home values, forecasts, new construction, rentals, and sales. They remain separate tables because their source schemas and meanings differ.

Modalities: Time-series

Source and scale

• Original dataset: misikoff/zillow-viewer
• Zillow source information: https://www.zillow.com/research/data/
• Processing source referenced by the original card: https://huggingface.co/datasets/misikoff/zillow
• Source revision: 83142d303d9ce00dd943d766d021cfd10a88b6d1
• Seven train products, 2,878,384 rows in total:

Config	Rows	Date range	Device TAG columns
days_on_market	586,714	2018-01-06 to 2024-02-03	region_id, region_type, home_type
for_sale_listings	578,653	2018-01-06 to 2024-01-06	region_id, region_type, home_type
home_values	117,912	2000-01-31 to 2024-01-31	region_id, region_type, home_type, bedroom_count
home_values_forecasts	31,854	2023-12-31 only	region_id, region_type
new_construction	49,487	2018-01-31 to 2023-11-30	region_id, region_type, home_type
rentals	1,258,740	2015-01-31 to 2023-12-31	region_id, region_type, home_type
sales	255,024	2008-02-02 to 2023-12-09	region_id, region_type, home_type

The rentals output may be emitted as multiple rentals_*.tsfile shards by the TsFile tool because it exceeds one million rows; all shards are one logical table.

TsFile schema

Every table has Time (INT64 milliseconds) and the TAG dimensions shown above. Shared source metadata such as size_rank, region, and state remains a FIELD when present. Product measurements remain numeric FIELDs; for example:

Config	FIELD measurements (normalized names)
days_on_market	mean_listings_price_cut_amount_smoothed, percent_listings_price_cut, mean_listings_price_cut_amount, percent_listings_price_cut_smoothed, median_days_on_pending_smoothed, median_days_on_pending
for_sale_listings	median_listing_price, median_listing_price_smoothed, new_listings, new_listings_smoothed, new_pending_smoothed, new_pending
home_values	bottom_tier_zhvi_smoothed_seasonally_adjusted, mid_tier_zhvi_smoothed_seasonally_adjusted, top_tier_zhvi_smoothed_seasonally_adjusted
home_values_forecasts	six month/quarter/year-over-year percentage fields, smoothed and unsmoothed
new_construction	median_sale_price, median_sale_price_per_sqft, sales_count
rentals	rent_smoothed, rent_smoothed_seasonally_adjusted
sales	all 11 source sales metrics, normalized to safe identifiers

region_type, home_type, and bedroom_count ClassLabel integer codes are decoded using the exact Hugging Face Parquet footer metadata for each file. Nullable measurements remain null.

Conversion notes

• Source Date timestamps are normalized to UTC epoch milliseconds in Time; the source Date column is not duplicated.
• Each product is sorted by its TAG columns and Time, with duplicate keys checked before schema-mode import.
• No source row or semantic measurement is dropped. Names containing spaces, punctuation, or parentheses are converted to deterministic lower-case safe identifiers.

Files and usage

The upload contains days_on_market.tsfile, for_sale_listings.tsfile, home_values.tsfile, home_values_forecasts.tsfile, new_construction.tsfile, the rentals_*.tsfile shards, and sales.tsfile.

from pathlib import Path
from tsfile import TsFileReader

path = Path("home_values.tsfile")
with TsFileReader(str(path)) as reader:
    table_name = next(iter(reader.get_all_table_schemas()))
    with reader.query_table(table_name, ["mid_tier_zhvi_smoothed_seasonally_adjusted"], batch_size=1024) as result:
        batch = result.read_arrow_batch()
        if batch is not None:
            print(batch.to_pandas().head())

License and attribution

The source dataset is released under the MIT license. See the original dataset card and Zillow Research for the source definitions.

Usage

Install the Apache TsFile Python SDK (pip install tsfile) and read a converted file:

from pathlib import Path
from tsfile import TsFileReader

path = Path("days_on_market.tsfile")
with TsFileReader(str(path)) as reader:
    schemas = reader.get_all_table_schemas()
    print("tables:", list(schemas))
    table_name = next(iter(schemas))
    table = schemas[table_name]
    columns = [column.get_column_name() for column in table.get_columns()]
    print("columns:", columns)
    field_names = [
        column.get_column_name()
        for column in table.get_columns()
        if column.get_column_name() not in {"Time", "time"}
    ]
    if field_names:
        with reader.query_table(table_name, field_names[:3], batch_size=1024) as result:
            batch = result.read_arrow_batch()
            if batch is not None:
                print(batch.to_pandas().head())