Adaptive batch sizes.

[KaspervanM]

Description

Now, the user has to input a batch size themselves, but often this is not the optimal batch size for the task.

Solution

Some processes take longer the higher the starting and ending values are. This is why there should be an algorithm that calculates the potential slope or curve that describes the duration of a subtask the further the task progresses. With this slope or curve, the best batch size can be assigned to the subtask.

The slope or curve could either be calculated before dividing the task into subtasks or during the process. The former would make the preprocessing take much longer, the latter would require a restructuring of the subtask system.