Using single sided inequality in nlpsol

[Sraobot]

I need to solve an optimisation problem using nlpsol with ipopt solver in Python. I have quite a few inequalities for that problem. For a double sided inequality like $a \leq g_1(x) \leq b$, I can use lbg and ubg by setting them with appropriate values. But, for a single sided inequality like $g_2(x) \leq 1$ or $g_2(x) \geq 1$ , what should be the right value for lbg or ubg when dealing with single sided inequalities.

Edit: I think, we can use ca.inf in the inequalities to make inequalities one-sided. I don't know, if there is a better way to deal with this.

[brunoml5]

You need to use ca.inf.

If you want a nicer interface, use opti.

[Sraobot]

Does opti also use ca.inf in the background or something else?

[brunoml5]

Yes

[Sraobot]

Ok. Thanks for the clarification.