Designing Experiments

On this page, you can find the algorithms to explore and optimize your reaction space.

1. Algorithm Selection

   Here, you can select whether you want to “Explore Reaction Space” using traditional Design of Experiments, or “Find Best Conditions” using Bayesian Optimization.

   Bayesian Optimization requires that you have already run some experiments.

2. Number of Experiments

   How many experiments should be designed. You can read about how to choose a reasonable number of experiments here.

3. Speed/Accuracy

   This setting adjusts the number of iterations the algorithm uses to find the best set of experiments. It is recommended to use “Balanced” or “Accurate”, especially for Finding Best Conditions or if using chemical descriptors.

4. Constrain Parameters

   In this table, you can constrain the values of the reaction parameters.

   For example, if you’ve speficied reaction parameter “Temperature”, but you want to run all reaction at 80 ˚C, you can add an entry with Name “Temperature” and Value “80”.

   In the Value column, it is possible to specify single or multiple values.

5. Design Experiments

   Once you are happy with the settings, you can start the algorithms by clicking this button.

   Designing the experiments takes from few seconds up to few minutes, depending on the size of the reaction space and the number of experiments.

Warning

For most reaction spaces, there are multiple optimal experiment desings. The algorithm might generate differrent sets of experiments each time - this is normal and expected behavior.