Mabe a stupid question but,
Is there a way of getting derivative of a profile histogram without applying a fit?
It depends what you mean by derivative.
Do you need more than two adjacent bins to compute the slope?
finding the slope using adjacent points, I guess, will give me big error at the end.
Is there a way like putting more dots between each adjacent points somehow ?An approximation method or smt like that?
Have a look at the spline classes.