Reproducibility Dataset for Engineering a Transition-Stage Digital Shadow for Resilient Road–Rail Freight Allocation under Shared Capacity and Delayed Information

Description

This dataset provides the complete reproducibility package for the study “Engineering a Transition-Stage Digital Shadow for Resilient Road–Rail Freight Allocation under Shared Capacity and Delayed Information.” The study develops a shared-capacity, rolling-horizon freight-allocation testbed for four stages of Iraq’s road–rail infrastructure transition: Existing, IREM-modernised, Phased-integration, and Planned-IDR. The package contains the model code, Iraqi corridor network inputs, asset–path incidence data, validation checks, scenario definitions, matched-seed simulation outputs, statistical analyses, capacity-sensitivity results, and figure-generation files. The principal factorial experiment comprises 10,800 runs combining four infrastructure states, three demand levels, five disruption conditions, six allocation controllers, and 30 matched stochastic replications. An additional 450 runs evaluate sensitivity to road, rail, and transfer-capacity assumptions. The six controllers are Static, Nominal-Rolling, Reactive-12h, Deterministic-DT, Chance-DT, and Perfect-Info. This design separates the value of rolling re-optimisation from the effects of updated observations, information delay, estimation error, and probabilistic capacity protection. The network representation combines an OpenStreetMap-derived current-road corridor with official railway and Development Road project information. Current operational capacities, disruption severities, variability, and observation-error parameters are transparent experimental assumptions. The results are reproducible simulation outputs and should not be interpreted as field-calibrated forecasts of Iraqi freight performance. The package includes a README with file descriptions, software requirements, reproduction instructions, evidence boundaries, citation information, and applicable licences.

Steps to reproduce

1. Download and extract `RIE_Data_Package.zip` to a local directory. 2. Install Python 3.11 or later. 3. Open a terminal or command prompt in the extracted package directory. 4. Create and activate a virtual Python environment if desired. 5. Install the required Python packages using: `pip install -r requirements.txt` 6. Change to the code directory: `cd code` 7. Run the principal factorial experiment and capacity-sensitivity analysis: `python iraq_transition_model_rie.py` This script reproduces: * 10,800 principal simulation runs. * 360 infrastructure–controller–demand–disruption summary cells. * 450 capacity-sensitivity runs. * Automated shared-capacity and execution-feasibility checks. 8. Run the statistical analysis: `python statistical_analysis_rie.py` This script reproduces the paired controller contrasts, bootstrap confidence intervals, paired t-tests, Cohen’s dz effect sizes, and Spearman sensitivity statistics. 9. Generate the final figures: `python make_rie_figures.py` The generated figures are saved in the `rie_figures` directory. 10. Compare the reproduced outputs with the archived files in the `results` and `figures` directories. The simulation uses fixed matched stochastic seeds to support deterministic reproduction of the demand, capacity, disruption, and observation-error scenarios. Minor differences in reported computational solve times may occur across operating systems, processors, Python versions, and SciPy HiGHS builds. These timing differences do not affect the simulated freight-allocation results.

Institutions

• University of Kufa

  An Najaf, Najaf

Categories

Licence