4. Performing the optimization of a kinetic model

To perform the optimization of a kinetic model:

1. Make sure you loaded the data

2. Select a model

   1. You must at least select a rate model or a vitrification model

   2. If a vitrification model is selected, you must provide a tg law

   3. If both rate and vitrification model, you must provide a coupling law

3. Provide the initial guess for your model

4. Switch to the ‘optimization method and parameters’ section

5. Select an optimization method

   1. You must at least select a global or local optimization method

   2. If you selected the ‘basinhopping’ or ‘shgo’ global optimization method, it is recommended to also select a local optimization method.

   3. You must select a cost function (rss_mean is recommended)

6. Start the optimization by clicking the ‘Start optimization’ button

⚠ The start of optimization can take a few seconds up to a few minutes to start, please be patient.

At the end of optimization, a result file is saved under ‘KinOpt/results/’.

Note

• The GUI will frequently be updated with the current state of optimization.

• An estimated remaining time is given. This estimated remaining time is not representative until at least ten optimization iterations have been performed.

• You’ll find a description of the kinetic modelds in Kinetic models module

• You’ll find a description of the optimization methods on Scipy documentation

• You’ll find a description of cost function in Optimization module