Online Resource for "Order-Induced Variance in the Moving-Range Sigma Estimator: A Total-Variance Decomposition"
Description
Reproduction code and reference outputs for the article "Order-Induced Variance in the Moving-Range Sigma Estimator: A Total-Variance Decomposition." The Python script Online_Resource_1.py reproduces every numerical result in the paper: the Normal-reference variance decomposition and adjacency-fraction values (Table 1, Section 4); the asymptotic efficiency accounting; the exact distribution-free component variances of Appendix B, cross-checked against the Normal closed forms and confirmed by Monte Carlo for the Uniform, Normal, t5, and Exponential distributions; and the Cryer and Ryan (1990) illustration with Figure 1. All computations use fixed random seeds, and console output matches the bundled expected_output.txt. The only empirical data are the 35 previously published chemical-process observations from Cryer and Ryan (1990, Table 2); all other quantities are computed or simulated. Requires Python 3.9+ with numpy and matplotlib. See README.md and RESULTS_MAP.md for details.
Files
Steps to reproduce
Install the dependencies (pip install -r requirements.txt; Python 3.9+, numpy, matplotlib), then run: python3 Online_Resource_1.py. The script prints all reproduced quantities to the console and writes Fig1.pdf, Fig1.eps, and Fig1.png to the working directory. To verify exact reproduction, redirect stdout to a file and diff it against the bundled expected_output.txt; with the fixed seeds (1 for the Monte Carlo confirmations, 123 for the permutation illustration) the console output matches exactly. Runtime is about 2-3 minutes, dominated by the 10^7 permutations used for Figure 1. Tested with Python 3.12, numpy 2.x, matplotlib 3.x.