numpy - How can I parallelize functions "leastsq" or/and "curve_fit" -

What is the best way to parallel the process appropriate for multicore computers using SCP functions? As far as I see from manual, there are no parameters like npocs in these functions, does it mean that they should not be parallel?

short answer: There is no built-in parallel in this filter. The list is in, but it is unclear whether it is worth it or not. (If you like to drive that code and report the result, it will be appreciated).

As it is said, you can parallel the evaluation of your objective function according to the similarity. For example, press it in a compiled extension (Cython, C, Fortran), release GIL and use OpenMP or explicit threads.