Code and data for: Annual effective dose from indoor radon-222 and thoron-220 in naturally ventilated equatorial dwellings: a first residential baseline for coastal Ecuador
Description
Raw track counts, derived activity concentrations and annual effective doses for 31 occupied dwellings measured with twin-cup LR-115 type II passive dosimeters over 90 days in the urban area of Daule, Guayas province, Ecuador, together with the scripts that reproduce every statistic and every figure of the associated article. Responses to the household questionnaire and the coordinates of the dwellings are withheld to preserve the anonymity of the participating families.
Files
Steps to reproduce
Requirements: Python 3.12 with NumPy 2.4, SciPy 1.17, pandas 2.2 and Matplotlib 3.10. No other dependency and no network access are needed. Download the four files of this record into the same working directory. Both scripts read Dataset_S1_Daule_indoor_radon_thoron.csv from that directory. Run: python Analysis_and_simulation.py The script first verifies the dose model, recomputing Eq. (6) of the article from the tabulated concentrations and stopping with an assertion error if any dwelling deviates by more than 0.001 mSv per year. It then prints the decomposition of the collective dose by parent isotope, the descriptive statistics of every radiometric variable, the comparison of the gamma, normal and log-normal fits by the Akaike information criterion, the exceedance probabilities predicted by each fitted model, the one-sample tests against the UNSCEAR, WHO and ICRP reference levels, the parametric and bootstrap confidence intervals, the correlation coefficients and the regression between the two isotopes, the lung-cancer risk indicators, the minimum detectable activity from Currie's criterion and the output of the diffusion-decay Monte Carlo model. Run: python Figures.py The script writes Figure_1.pdf to Figure_7.pdf in vector format, together with PNG previews at 600 dpi. These are the figures published with the article. Reproducibility: the bootstrap and the Monte Carlo model use a fixed seed (2026), so every reported number and every figure is identical from run to run. The figures regenerated from this record match the submitted files byte for byte. Note on Figure 6: the group means shown there come from the household questionnaire, which is not published in order to preserve the anonymity of the participating families. Those group statistics, reported in Table 6 of the article, are therefore hard-coded in Figures.py. Expected runtime: under one minute on a standard laptop.
Institutions
- Universidad Técnica de ManabíManabí, Portoviejo