Reproducibility package: predictive coverage audit

Input data: I-Cheng Yeh (2018), Real Estate Valuation, UCI Machine Learning
Repository. https://doi.org/10.24432/C5J30W
Dataset licence: Creative Commons Attribution 4.0 International.
https://creativecommons.org/licenses/by/4.0/
The CSV is a format conversion of the original UCI spreadsheet. Row values
and column order are retained. Data describe Taiwan, not the United States.

Run: python reproduce.py
Requires Python 3, numpy and pandas. No API key or network access is needed.
The script writes results.json beside itself and prints summary metrics.
The original downloaded ZIP SHA256 is recorded in results.json.

The analysis code is supplied under the MIT licence:
Copyright (c) 2026 SHLAEN RETAIL LLC.
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files, to deal in the Software
without restriction, including without limitation the rights to use, copy,
modify, merge, publish, distribute, sublicense, and/or sell copies, and to
permit persons to whom the Software is furnished to do so, subject to the
following conditions: this copyright notice and permission notice shall be
included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO MERCHANTABILITY, FITNESS FOR A
PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR
IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Important limitations: date ties cross the chronological split boundary;
this is not a strictly out-of-time evaluation. Repeated random splits share
observations. Marginal coverage does not guarantee subgroup or conditional
coverage. Geographic proximity and repeated properties cannot be ruled out
with this dataset. No untested model comparison or causal claim is supported.
