What is Calculation fingerprint?
A “fingerprint” of a calculation, made from everything that determines its result: inputs, model, mechanism, solver settings and software versions.
An identical calculation never needs to be solved twice. If fingerprints match, the stored result can be safely reused.
Just as matching fingerprints identify the same person, matching calculation fingerprints identify the same calculation.
Shown in each result's provenance. Changing display texts or labels never changes a fingerprint.
The block type, normalized inputs (SI units, normalized compositions, defaults filled in), the contents of incoming streams, the mechanism content hash, the physics model version, solver settings and the Cantera and Laplaxian versions are hashed together with SHA-256.