Better results format

[j-brendel]

As output of the smooth simulation a format is nedded that fulfils certain criteria:

• Readibility: The results should be easily readible, also for persons, that don't use smooth (like .csv/.json...)
• Reproducibility: The results should contain all information needed to reproduce them: This includes
  • The simulated model
  • The SMOOTH version (commit-tag?) used to produce the results
  • more?
• Completeness: All relevant results should be accessible. In the .pickle files created now, e.g. the external components costs are not stored

[j-brendel]

One idea would be to

• safe the dictionaries (results and model) into a .json file
• also use the .json file for input

idea of the .json file:

It should be only one file, that can be used for simulation and for optimization.

for input:
{
model: {
'busses': busses,
'components': components,
'sim_params': sim_params,
}
}

• remember optimzation params

for output:
{
model: {
'busses': busses,
'components': components,
'sim_params': sim_params,
},
results: {
chosen_model{}
}
}

Beware: Lambda functions in model definition on optimization -> talk to stefan