Lower bound convergence issue

[harshangrjn]

This may be related to issue #134. But reporting it just so that it can be fixed.

@variable(m, 2 <= h[1:2] <= 3)
@NLobjective(m, Min, 2*(h[1])^4.0)
@constraint(m, h[1] + h[2] >= 0.5)

For the above problem, Alpine's lower bounding iterations get stuck at 8.0, while the global optimum is 32.0. This issue doesn't appear when the co-efficient 2 is not in the objective or when the lower bound on the variable h is equal to 1.0. Most probably an issue with obj. expression parsing.

[harshangrjn]

Here is the updated output:

====================================================================================================
| Iter   | Incumbent       | Best Incumbent      | Lower Bound        | Gap (%)         | Time      
| finish | 32.0            | 32.0                | 32.0               | 0.0             | 0.0s             
====================================================================================================