masson-group-ohiouniversity/Equilibrist: Equilibrist 2.0.1

A newer version is available. Equilibrist 3.0 beta adds tabs with global fits across techniques, isothermal titration calorimetry, supercomputer offload and tutorials 19–24: https://github.com/masson-group-ohiouniversity/Equilibrist/releases/tag/v3.0-beta (the current code is on the main branch). This 2.0.1 page is kept as distributed. Equilibrist 2.0.1 A kinetics fit could report a perfect fit without having fitted anything. If you ran one under 2.0, please read this. When a data column was named after a $variables expression — say %GH = 100 * GH/(G+GH) — Equilibrist looked for that column among the simulated species, didn't find it, and formed no residuals. With nothing to minimize, the optimizer stopped where it started. The output was R² = 1.000, RMSE = 0, and "Kinetics fit complete", with the rate constants sitting at whatever you had typed as initial guesses. Bootstrap CIs on the same fits came back empty. Also in this release: a fit whose columns match no species and no variable now fails with a message naming what it found and what it expected. A typo in a column header used to produce R² = 1.000 as well. R² and RMSE now return NaN when undefined rather than 1.0 and 0.0. Equilibrium fitting, NMR and spectral kinetics, and kinetics fits whose columns name species directly are unchanged. Sessions saved under 2.0 still load.