Extra Notes

Extra Notes

• Mandatory top-level keys: mess_inputs, ct_yaml, experiments.
• Exactly one of initial_ratio or initial_concentration must be set for each experiment.
• Experiment pressure units are canonicalized and converted to Pa for cantera simulation; MESS pressure grid is built in bar.
• Ensembles (generations, GOAT, Nelder-Mead Swarm) are referenced by a tag name (e.g., G0001, GT-1, etc.) and must exist in the database for postprocessing to succeed.
• Postprocessing writes extrapolated results back into the primary run databases (KMO_DB_KIN / KMO_DB_SIM), into the same per-generation tables where the models were created; it no longer creates separate PP_DB_KIN.db / PP_DB_SIM.db files or X-prefixed tables. Already-computed rate coefficients are reused (only missing (P, T) conditions are recomputed), and extrapolated simulations are appended with banded experiment ids so the original run's results are never overwritten. This change is forward-only: pre-existing PP_DB_*.db files and old X-prefixed tables are not migrated.
• Inputs can be prepared using the command kmo_start, which will generate a JSON file with the correct structure and chosen values. The JSON can then be edited manually or reloaded into the GUI for further editing.
• While some analysis is possible using the GUI (called with keyword kmoui), it is still in development, and analysis of the results is best done using Jupyter notebooks or scripts that load the database and perform postprocessing.