Data for "Exhaustive Symbolic Integration: Integration by Differentiation and the Landscape of Symbolic Integrability", Desmond (2026)

Data upload package for the paper "Exhaustive Symbolic Integration: Integration by Differentiation and theLandscape of Symbolic Integrability".

Contents

- `results/`: final paper result artifacts, CAS-comparison JSON files, rho and decomposition summaries, lookup-equivalence pickle files, raw derivative and fingerprint outputs, and `esi_hash_index.json`.- `function_catalogues_unique_equations.tar.gz`: the five-basis ESR function catalogues needed to reproduce the pipeline inputs. The archive contains only `unique_equations_k.txt` files, not the much larger ESR tree, match, inverse substitution, or generation-intermediate files.- `MANIFEST.tsv`: file list with byte sizes and broad categories.- `CHECKSUMS.sha256`: SHA256 checksums for every upload file.- `zenodo_metadata.json`: draft metadata fields for the Zenodo record.

Scope

The package is intended to provide the data files needed to:

1. run `esi_integrate.py` and `esi_integrate_demo.ipynb` using the lookup pickle files and hash index;2. regenerate the paper tables and figures from the stored summaries;3. audit the final CAS-resistance claims from the saved CAS cascade and stress test artifacts;4. rerun the ESI pipeline from the included unique-equation catalogues.

The code, paper source, lookup script, notebook, and analysis scripts are maintained in the GitHub repository:`https://github.com/harrydesmond/ExhaustiveSymbolicIntegration`.

It intentionally does not include raw ESR generation intermediates such as `trees_k.txt`, `orig_trees_k.txt`, `matches_k.txt`, `inv_subs_k.txt`, or local cluster logs. Those files are many GB, are reproducible from ESR plus the grammar definitions, and are not used directly by the paper analysis scripts. CAS executables, Mathematica/RUBI installations, FriCAS/Sage containers, and Slurm outputs are also excluded from this data record; use the GitHub repository for code.