Daylighting design and post-construction assessment of a classroom: validated Radiance model, illuminance measurements and reproduction scripts

Companion dataset of the article "Daylighting design and post-construction assessment of a classroom: illuminance measurements and simulations" (G. Barrios, J. M. Ochoa, I. Marincic, C. Galindo, G. Huelsz; submitted, 2026). Minimal, self-contained package to reproduce the validated Radiance two-phase (daylight-coefficient) simulation of a naturally lit classroom at the Instituto de Energías Renovables, UNAM, in Temixco, Mexico (18.85° N, 99.14° W), and to compare it with luxmeter measurements taken on June 26 and November 20, 2024 on a 7 × 9 grid (63 points, 1.08 m spacing) at 0.75 m above the floor, hourly from 9:00 to 17:00. Contents:- model/: Radiance geometry of the classroom with the sliding windows open as during the measurements, 36 school chairs and the surrounding terrain; calibrated materials (glazing transmittance τ = 0.80 as the only calibrated parameter, measured floor, wall and ground reflectances); sky receiver and sensor positions.- data/experimental/: raw luxmeter CSV files, one per hour and day (kilolux).- data/weather/: annual WEA file for Temixco with the sky conditions of the 18 measured hours replaced by 1-min ESOLMET station means, plus the table of injected values.- scripts/run_simulation.sh: regenerates the annual illuminance file (gendaymtx, oconv, rfluxmtx, dctimestep/rmtxop) in about one minute.- notebooks/illuminance_maps.ipynb: plots measured versus simulated illuminance maps for both days and prints the validation metrics (NMBE, CV(RMSE), NMAE, R², GOF).- scripts/clean.sh, pyproject.toml, uv.lock: cleanup and reproducible Python environment (uv). Simulation outputs (octree, sky matrix, daylight-coefficient matrix, .ill file) are not included; they are regenerated by the script. Requirements: Radiance ≥ 5.4 (tested with 6.0.1) and uv. See README.md for the full description of the model, the data format and the measurement protocol.