Hi, I want to convexify this function f(x,y)=log((x+a)/b)+log((y+a)/b) in CVX. Is there any idea on how this can be done? The constraints are like log(1+(x/a))>=c, log(1+(y/a))>=d. Here, a, b, c, d are constants. I appreciate any feedback you can share.

Those are already convex constraints (of the form {concave} >= constant). So they should be enterable “as is” into CVX.

Thank you for your reply. What about the objective function f(x,y)=log((x+a)/b)+log((y+a)/b)?

Presuming `b`

is a constant, that is concave, so it can be maximized in CVX. If you want to minimize it, CVX is the wrong tool (you could try YALMIP).

Thank you again for your reply. It would help me a lot.