I wrote a tutorial on how to do linear and non linear fit in Python with scipy. I give practical examples that can be applied very easily to several use cases. In particular, I think that People will find interesting the discussion and the code example on how to find ideal intial Parameters for a non linear fit.
In particular I give an example on how to use the functions curve_fit() and differential_evolution() to do a (kind of) grid search for the itinial Parameters that are as close to the right one as possible. As usual, I Keep all my code on this github respository. So feel free to check the Jupyter Notebook directly, or download code and data files.
Jupyter Notebook: LINK
You can find more projects I worked on on my blog https://udata.science.
Commets and Feedback is, as usual, welcome.
Umberto Michelucci
18 July 2017, Zürich.