Refactor single and dynamic mode into one named observation
The idea of this refactoring would be to only have three true running modes: observation, parametric and calibration, so not dynamic and single as separate running modes. Now each of the modes those would have an option to specify a dynamic section in configuration file like so:
observation:
dynamic:
times: np.linspace()
outputs:
...
The outputs of single and dynamic pipeline should be in the same format - xarray dataset.
Further down the road the dynamic flag should be possible for all running modes (with some more development):
observation/parametric/calibration:
dynamic:
times: np.linspace()
...