Reproduce results from: Hessen (2025); Table 1, Table 2, and Figure 2 (Q7260362)

A FAIR-compliant **Snakemake** workflow that reproduces all main results of Hessen (2025) by fitting a convexity-constrained Random Effects Graded Partial Credit Model (RE-GPCM) to the Dekovic child aggression dataset (5 items, 4 response categories). The model is optimised via BFGS with analytic gradients using the **R** package **calculus**, with the convexity constraint enforced by reparametrising threshold parameters as cumulative sums of positive increments. All R package versions are pinned in a renv lockfile included in the RO-Crate, ensuring numerical results match those reported in the paper.\N\N### Main steps\N\N**fit_re_gpcm** — loads Dekovic_Aggr.dat, runs BFGS optimisation with configurable seed and v parameter, outputs table1.csv, table2.csv, figure2.png, and an intermediate params CSV.\N\N**format_tables** — renders table1 and table2 as publication-quality PNG images using a Python formatting script.