I have some simulation data for a discrete LTI system which are provided as unit step response models between the 5 inputs (rows) and 12 outputs (columns).

I'd like to convert the model to a state space representation in python-control but I'm not sure how to do that.

Based on what I've read so far, it seems like the best way is to convert the step response models to impulse response first, then use control.era() to get a state space representation. Am I on the right track?

Thanks,
Siang