I have two questions. If anybody helps me, I shall be highly obliged. Thanks in advance.

(i) Can anybody provide me the generalized primal version (i.e. maximization problem) of input-oriented VRS model?

I have tried it myself with an example. Actually, I tried to convert an input-oriented VRS model from dual version (i.e. minimization problem) to primal version (i.e. maximization problem). I found in case of objective function and constraints, an additional variable will come that does not exist in case of CRS model primal form (maximization problem). Additionally, the denominator part that is equal to 1 (constraint) in CRS primal form, it becomes less than or equal to 1 in VRS primal form. Am I right?

(ii) I know that in input-oriented VRS model, pure technical efficiency is calculated, scale efficiency is eliminated. I also know that in dual form it is done by adding a constraint summation of lambda or dual weights is equal to 1. However, I cannot relate these things.

Additionally, how this (i.e. pure technical efficiency without scale efficiency) can be explained with the primal form of VRS model?

Any kind of suggestion will help me a lot.