MIT Adaptive Update law construction in Simulink

I want to track a desired trajectory of 2 dof planar manipulator. I want to apply direct adaptive control to track the trajectory. The adaptive law looks like below:

thetadot= gamma*error*(delta(e)/delta(theta))

where theta= adpative varibale. gamma=adaptation gain error=(desires pos-actual pos ) of robot joint.

rate of change of theta (thetadot) is very slow.

My question is how do I construct (delta(e)/delta(theta)) in simulink. Can it be expressed as (e(k)-e(k-1))/(theta(k)-theta(k-1))? If possible, please instruct me how to develop it in simulink?

Thank you.