Currently, the optimize.response_fit function does not use the measurement (y-value) error or standard deviation in the curve fitting. However, it would probably be good to allow this by using something like Chi-squared metric instead of SSE when the error values are passed in to the function. The function has the argument to accept the error values but does not actually use it now.
Currently, the
optimize.response_fitfunction does not use the measurement (y-value) error or standard deviation in the curve fitting. However, it would probably be good to allow this by using something like Chi-squared metric instead of SSE when the error values are passed in to the function. The function has the argument to accept the error values but does not actually use it now.