CA final costing - learning curve

Abhishek (Student) (78 Points)

15 January 2019  
Using the formulae y=a*x^b (where y is average time per unit taken to produce x units; a is time taken to produce the first unit; x is the number for units for which the average time(I.e., y) is required to be found out; b is the learning co-efficient, I.e., B=log(learning rate)/log(2) ) implies that the learning is for a unit series of 1, 2, 4, 8, 16 and so on. That is to say if in the formulae, x is taken as say 5 or 20 it is from in between the above unit series.
But is there a formulae or adjustment for a unit series of say 5, 10, 20, 40, 80 and so on or say 120, 240, 480, 960 and so on?