Estimate = false doesn't work for estbounds

[a-torgovitsky]

r <- estbounds(data, lpm, solver = "gurobi", norm = 1, kappa = .1,
               estimate = FALSE) 

seems to exit ok, but then I get

> str(r)
List of 6
 $ ub    : num(0) 
 $ lb    : num(0) 
 $ est   : logi FALSE
 $ call  : language estbounds(data = data, lpmodel = lpm, kappa = 0.1, norm = 1, solver = "gurobi", estimate = FALSE)
 $ norm  : num 1
 $ solver: chr "gurobi"
 - attr(*, "class")= chr "estbounds"

So apparently nothing happened

[conroylau]

Done! I have fixed the bug that caused the problem here, which prevented the module from estimating the bounds although the original problem is infeasible.

Since the original problem is infeasible, the above code will give the estimated bounds as follows:

> summary(r)
Estimated bounds: [0.05273, 0.47834] 
Norm used: 1 
Solver: gurobi 

Thanks!