Data for: Validating AI-Driven Decision-Making in Renewable Energy Systems
Description
Supporting data for the review "Validating AI-Driven Decision-Making in Renewable Energy Systems" (Renewable and Sustainable Energy Reviews, submitted). Contains the coded corpus of 1,591 screened records (1,271 primary studies) on the seven-level Decision-Readiness scale; the codebook (DR decision tree); the PRISMA search log; the double-coding, reconciliation and recheck files; the extraction template and value-extraction data; and the Python scripts that reproduce every figure and count in the paper.
Files
Steps to reproduce
Requirements: Python 3 with matplotlib and numpy (tested with Python 3.14, matplotlib 3.11, numpy 2.5). Download the dataset. Place the scripts folder and all CSV files from the data folder in the same top-level folder, i.e. folder/scripts/ and folder/*.csv. From that top-level folder, run in this order: py scripts/make_schematics.py py scripts/make_priority_figures.py py scripts/make_figures.py py scripts/make_figures2.py py scripts/make_checklist.py Figures are written to figures/pdf (vector), figures/png and figures/svg. Every count is computed from the CSVs at run time, except the cell counts in Figure 9, which are entered in make_figures.py from B8_Q3_sim_to_real.csv. File name to article figure: fig10_review_growth = Fig. 1; fig02_dr_scale = Fig. 2; fig01_prisma = Fig. 3; fig06_composition = Fig. 4; fig04_distribution = Fig. 5; fig05_retention = Fig. 6; fig03_temporal = Fig. 7; fig11_uncertainty_carried = Fig. 8; fig09_transfer_evidence = Fig. 9; fig12_failure_heatmap = Fig. 10; fig13_reporting_checklist = Fig. 11; fig08_failure_gradient = Supplementary Fig. S1. The DR codes can be checked by hand against the codebook (DR_DECISION_TREE.pdf). check_compliance.py, word_budget.py, check_reduction.py and make_sankey.py were used to check the manuscript during drafting and need the LaTeX source; they are not needed to reproduce the results.