Data recomposition
Summary
As a user, I want to create a map from predictions made by neurisp model, so I can use the map for other applications or use it a as another input for the model.
Description
As explained in issue #8 (closed) the input of the model are puntual values, also the model produces puntual values. For this reason is necessary to write an algorithm that recompose a map from them.
It is also required to implement a mechanism to update the dataset for train/inference with the predicted map.
List of steps needed by user story:
- create an emtpy 2d array with the same size of the input
- take output calculated from model and place it inside the emtpy array
- read crs and no value from input stack
- set crs and no value to 2d array and save it as a map