LINDO-like sensitivity analysis report from gurobi shell

[Betty Love]

Is there a way to get gurobi to output a LINDO-like sensitivity analysis report from the gurobi shell? I'd like ranges on each cost coefficient and rhs value for which the optimal basis remains optimal.

Thanks!

[Angelo]

Hello Betty,

I am using Gurobi through AMPL and I have added some appropriate commands in a script file, in order to output a LINDO-like sensitivity analysis report. Please find attached an example model (.mod) file and a script (.run) file. Although, I am not using the Python shell, I believe/hope that these might be helpful to you. Perhaps, you can try something similar with the Python syntax. The sensitivity analysis report, for the attached example, is as follows:

  Variable     Value   Reduced Cost

-----------------------------------

: _varname[j] _var[j] _var[j].rc    :=

1   'x[1]'       3         0

2   'x[2]'       1.5       0

;

   Row       Slack or Surplus   Dual Price 

------------------------------------------

: _conname[j] _con[j].slack _con[j].dual    :=

1   R1              0            0.75

2   R2              0            0.5

3   R3              2.5          0

4   R4              0.5          0

;

Ranges for the objective-function coefficient of the variable:

---------------------------------------------------------------

: _varname[j] _var[j].sensobjlo _var[j].sensobjhi    :=

1   'x[1]'          2                    6

2   'x[2]'          3.33333             10

;

Ranges for the RHS of the constraints:

--------------------------------------

: _conname[j] _con[j].sensrhslo _con[j].sensrhshi    :=

1   R1               20                36

2   R2                4                 6.66667

3   R3               -1.5          1e+100

4   R4                1.5          1e+100

;

As you can see, they are pretty similar to LINDO...

Best regards,

Angelo