CoolFace
Datasetpublic

covaga/Philippines-Real-Estate-Asking-Prices

Philippines Real Estate Asking Prices 241,496 de-identified property listings with asking prices in Philippine pesos (₱), covering condos, houses, lots and commercial property across 17 regions. This is a tabular housing-market dataset for the Philippines — useful for price-per-square-meter analysis, city comparisons, geospatial EDA, affordability research, and regression baselines. These are asking prices from public listings, not closed / transaction / sold prices. They are… See the full description on the dataset page: https://huggingface.co/datasets/covaga/Philippines-Real-Estate-Asking-Prices.

sourceHugging Faceotherupdated 27d agoView on Hugging Face
0likes77downloads
Dataset Card

Philippines Real Estate Asking Prices

241,496 de-identified property listings with asking prices in Philippine pesos (₱), covering condos, houses, lots and commercial property across 17 regions.

This is a tabular housing-market dataset for the Philippines — useful for price-per-square-meter analysis, city comparisons, geospatial EDA, affordability research, and regression baselines.

These are asking prices from public listings, not closed / transaction / sold prices. They are not an official index (not BSP RPPI, not BIR zonal values).

Snapshot: 2026-08-30. Observation window: 2026-02-22 → 2026-08-30.

Companion city-level aggregates (medians, inventory, quality): covaga/philippines-real-estate-market-prices on Hugging Face.

Why it exists

Structured Philippine real-estate microdata is scarce compared with US/EU housing datasets. Official series (BSP Residential Property Price Index, BIR zonal values) are aggregates. This file is listing-level asking prices, stripped of portal, agent and photo identifiers so it can be shared for research and learning.

What's in the files

fileformatrows
listings.parquetParquet (zstd)241,496
listings.csv.gzgzip CSV (same schema)241,496
splitn
All listings241,496
Still active at snapshot49,884
Condo88,132
House97,133
Lot49,039
Commercial7,192
With lat/lng240,202
With ₱/m²207,539
Distinct city strings1,530
Provinces84
Regions17
Opaque sources (S01–S09)9

All rows are listing_type = sale.

Schema

columntypenotes
listing_idstringStable opaque hash. Not a portal id.
source_idstringOpaque portal label S01…S09. Same value = same source.
canonical_idstringOpaque group id when the pipeline thought two rows were the same property.
listing_typestringsale only in this release.
property_typestringcondo / house / lot / commercial
property_subtypestringFiner type when known (studio, townhouse, house_and_lot, …). Often null.
price_phpfloatAsking price in PHP.
sqmfloatFloor area, m².
lot_area_m2floatLot/land area, m², when known.
price_per_sqmfloatprice_php / sqm. Null if sqm missing or ₱/m² outside 1,000–1,000,000.
bedroomsint
bathroomsint
citystringRequired. Parsed locality; spelling is not fully standardized (1,530 distinct values).
districtstring
barangaystring
provincestring
regionstring
lat / lngfloatRounded to 3 decimal degrees (~100 m).
n_photosintPhoto count only. No image URLs.
first_seen / last_seendateUTC dates the pipeline first/last observed the listing.
is_activeboolStill listed at snapshot.
quality_scoreintPipeline score 0–100. Null if never scored.
new_build_flagboolHeuristic: pre-selling / brand-new / RFO.

What was removed (on purpose)

  • —Portal names and listing URLs
  • —Titles and descriptions (copyright + incidental PII)
  • —Photo URLs (only n_photos)
  • —Agent / seller identity

Cleaning already applied

  • —Dropped missing/invalid price, missing city, unknown property type
  • —Dropped pipeline needs_review outliers
  • —Asking price capped at ₱1,000,000,000
  • —₱/m² recalculated; out-of-range values set to null
  • —Coordinates rounded

city names are still messy. That is the main remaining data-quality task if you want clean city-level stats.

Caveats

  • —Asking ≠ sold. Listings can sit, get discounted, or never transact.
  • —Coverage is whatever public portals exposed to the pipeline in this window — not a census of the housing stock.
  • —city is a parsed string, not a guaranteed PSGC code.
  • —Do not re-identify portals, agents, or exact properties.

License

CC BY 4.0. Research and analysis. Attribution appreciated. No warranty of completeness.

Citation

Christian (covaga.xyz), Philippines Real Estate Asking Prices , snapshot 2026-08-30.