Description
💡The new DataLab core library (see #166) could be an opportunity to reconsider the design of the DataLab processing functions.
Currently, the processing functions working on signal and image objects rely on guidata.dataset.DataSet objects for input parameters. This is very convenient for the developer because it allows to create a GUI for the processing functions automatically, but it is not very flexible for the user because it forces to instantiate a DataSet object with the right parameters before calling the processing function (this can be cumbersome especially when dealing with simple processing functions requiring only a few parameters).
Thus, it could be interesting to consider a more flexible and simple design, where the processing parameters would be passed as keyword arguments to the processing functions. The DataSet objects could be handled internally by the processing functions (e.g. by calling the DataSet.create method with the keyword arguments passed by the user). This would allow to keep the automatic GUI generation feature for the processing functions, but would also allow to call the processing functions directly with keyword arguments, without having to create a DataSet object first.