spb/wp2_uqo Public
UQO Working Paper No. 2 — Decoding Real Estate Descriptions: text-based hedonic analysis of housing listings.
TeX 73.8%
Python 26%
1<!-- Author: Simon-Pierre Boucher — contact@spboucher.ai -->23# data/ — not tracked by git45This directory is excluded from version control because:671. `raw/louka.db` (268 MB) exceeds GitHub's 100 MB per-file limit;82. the database contains a leftover application `users` table (credentials) that9 must not be published;103. the listing texts are proprietary platform content (Realtor.ca / Centris).1112## Restoring the data locally1314Place the SQLite database at:1516```17data/raw/louka.db # table `properties`: 46,479 listings, 5 categories18```1920(or set the environment variable `WP2_DB_PATH` to its location), then run the21pipeline from the repository root:2223```bash24python3 scripts/01_prepare_data.py25python3 scripts/02_similarities.py26python3 scripts/03_models.py27python3 scripts/04_robustness.py28python3 scripts/05_figures.py29python3 scripts/06_tables.py30```3132The pipeline recreates everything under `data/processed/`:3334| File | Content |35|---|---|36| `houses.parquet` | analysis sample (17,087 × 14, incl. remarks text) |37| `embeddings_maisons.npy` | 17,087 × 384 sentence embeddings (cache) |38| `sim_matrix_maisons.npy` | 17,087 × 20 cosine-similarity matrix |39| `hedonic_maison_results.csv` | analysis dataset used by all models/figures |40| `legacy/` | pre-refactor artifacts kept for byte-level comparison |4142Embeddings are deterministic for a given `all-MiniLM-L6-v2` model version, so a43fresh run reproduces the committed results exactly (see `CHANGES.md` §3).44