Lp Dual Shadow
1. **Problem Statement:** We have a primal linear program (P):
$$\max w = 5x_1 + 8x_2$$
subject to
$$x_1 + x_2 \leq 6$$
$$x_1 + 2x_2 \leq 8$$
$$x_1, x_2 \geq 0$$
The final simplex tableau for (P) is given as:
$$\begin{array}{c|cccc|c}
x_1 & 1 & 0 & 2 & -1 & 4 \\
x_2 & 0 & 1 & -1 & 1 & 2 \\
\text{z-row} & 0 & ? & 2 & 3 & 36 \\
\end{array}$$
Problem (D) is the dual of (P).
---
2. **(a) Phases required to solve (D):**
- The dual (D) has constraints corresponding to variables in (P).
- Since (P) has \(\leq\) constraints with slack variables, the dual will have \(\geq\) constraints or equalities.
- The starting tableau of (D) will have artificial variables if constraints are equalities or \(\geq\).
- Here, (D) has two constraints (dual variables \(u_1, u_2\)) and two slack variables in (P).
- Because (D) has two artificial variables (due to \(\geq\) constraints), solving (D) requires **phases 1 and 2**.
**Answer:** phases 1 and 2, because the starting tableau of (D) has two artificial variables.
---
3. **(b) Shadow price of 1st constraint of (P):**
- Shadow prices correspond to the values of dual variables at optimum.
- From the final tableau, the coefficients of slack variables \(y_1, y_2\) in the objective row give shadow prices.
- The coefficient of \(y_1\) in the z-row is 2.
**Answer:** 2
---
4. **(c) Shadow price of 1st constraint of (D):**
- The shadow price of the 1st constraint of (D) corresponds to the value of \(x_1\) in the final tableau.
- From the tableau, the value of \(x_1\) in the solution column is 4.
**Answer:** 4
---
5. **(d) Value of \(u_2\) in optimum solution of (D):**
- The dual variables correspond to slack variables in (P).
- From the tableau, the value of \(y_2\) (dual slack) in the solution column is 2.
- Hence, \(u_2 = 2\).
**Answer:** 2
---
6. **(e) Interpretation of '?' (reduced cost \(r_{x_2}\)) in final tableau:**
- The reduced cost \(r_{x_2}\) relates to the slack of the 2nd constraint in (D).
- The shadow price of \(x_2\) is the value of \(u_2\) which is 2.
- Complementary slackness implies if shadow price is positive, slack must be zero.
- Therefore, \(r_{x_2} = 0\).
**Answer:** \(r_x = 0\)
---
**Summary of answers:**
(a) phases 1 and 2, because the starting tableau of (D) has two artificial variables
(b) 2
(c) 4
(d) 2
(e) \(r_x = 0\)