OCR D1 2008 January — Question 6 13 marks

Exam BoardOCR
ModuleD1 (Decision Mathematics 1)
Year2008
SessionJanuary
Marks13
PaperDownload PDF ↗
Mark schemeDownload PDF ↗
TopicThe Simplex Algorithm
TypePerform one Simplex iteration
DifficultyModerate -0.3 This is a standard Simplex algorithm question testing routine procedural knowledge: setting up a tableau, identifying pivot elements using standard rules, performing one iteration with row operations, and recognizing when the algorithm fails (negative z-value). While multi-step with 13 marks total, each part follows textbook procedures with no novel problem-solving required, making it slightly easier than average.
Spec7.07a Simplex tableau: initial setup in standard format7.07b Simplex iterations: pivot choice and row operations

  1. Represent the linear programming problem below by an initial Simplex tableau. [2] Maximise \quad \(P = 25x + 14y - 32z\), subject to \quad \(6x - 4y + 3z \leqslant 24\), \qquad\qquad\quad \(5x - 3y + 10z \leqslant 15\), and \qquad\qquad \(x \geqslant 0\), \(y \geqslant 0\), \(z \geqslant 0\).
  2. Explain how you know that the first iteration will use a pivot from the \(x\) column. Show the calculations used to find the pivot element. [3]
  3. Perform one iteration of the Simplex algorithm. Show how each row was calculated and write down the values of \(x\), \(y\), \(z\) and \(P\) that result from this iteration. [7]
  4. Explain why the Simplex algorithm cannot be used to find the optimal value of \(P\) for this problem. [1]

(i)
AnswerMarks Guidance
Px y
1-25 -14
06 -4
05 -3
Rows and columns may be in any order; Objective row with -25, -14, -32B1
Constraint rows correct (condone omission of \(P\) column)B1
[2]
(ii)
AnswerMarks
\(x\) column has a negative value in objective rowB1
'negative in top row', '-25', or similar; 'most negative in top row' ⇒ bod B1B1
Cannot use \(y\) column since it has negative entries in all the other rowsB1
Correct reason for not choosing \(y\) columnB1
\(24 \div 6 = 4\); \(15 \div 5 = 3\); Least non-negative ratio is 3, so pivot on 5B1
Both divisions seen and correct choice made (or both divisions seen and correct choice implied from pivoting)B1
[3]
(iii)
AnswerMarks Guidance
10 -29
00 -0.4
01 -0.6
Follow through their sensible tableau (with two slack variable columns) and pivotM1
Pivot row correct (no numerical errors); Other rows correct (no numerical errors)A1
New row 3 = \(\frac{1}{5}\) row 3B1
New row 1 = row 1 + 25×new row 3 o.e.B1
New row 2 = row 2 - 6×new row 3 o.e.B1
Calculation for pivot row; Calculation for objective row; Calculation for other rowB1
\(x = 3\), \(y = 0\), \(z = 0\)B1 ft
\(P = 75\)B1 ft
\(x\), \(y\) and \(z\) from their tableau; \(P\) from their tableau, provided \(P \geq 0\)B1
[2]
(iv)
AnswerMarks
Problem is unbounded; No limit to how big \(y\) (and hence \(P\)) can be; Only negative in objective row is \(y\) column, but all entries in this column are negativeB1
Any one of these, or equivalent. If described in terms of pivot choices, must be complete and convincingB1
[1]
## (i)
| P | x | y | z | s | t |
|---|---|---|---|---|---|
| 1 | -25 | -14 | 32 | 0 | 0 |
| 0 | 6 | -4 | 3 | 1 | 0 | 24 |
| 0 | 5 | -3 | 10 | 0 | 1 | 15 |

Rows and columns may be in any order; Objective row with -25, -14, -32 | B1 |
Constraint rows correct (condone omission of $P$ column) | B1 |
| [2] |

## (ii)
$x$ column has a negative value in objective row | B1 |
'negative in top row', '-25', or similar; 'most negative in top row' ⇒ bod B1 | B1 |
Cannot use $y$ column since it has negative entries in all the other rows | B1 |
Correct reason for not choosing $y$ column | B1 |
$24 \div 6 = 4$; $15 \div 5 = 3$; Least non-negative ratio is 3, so pivot on 5 | B1 |
Both divisions seen and correct choice made (or both divisions seen and correct choice implied from pivoting) | B1 |
| [3] |

## (iii)
| | 1 | 0 | -29 | 82 | 0 | 5 | 75 |
| | 0 | 0 | -0.4 | -9 | 1 | -1.2 | 6 |
| | 0 | 1 | -0.6 | 2 | 0 | 0.2 | 3 |

Follow through their sensible tableau (with two slack variable columns) and pivot | M1 |
Pivot row correct (no numerical errors); Other rows correct (no numerical errors) | A1 |
New row 3 = $\frac{1}{5}$ row 3 | B1 |
New row 1 = row 1 + 25×new row 3 o.e. | B1 |
New row 2 = row 2 - 6×new row 3 o.e. | B1 |
Calculation for pivot row; Calculation for objective row; Calculation for other row | B1 |
$x = 3$, $y = 0$, $z = 0$ | B1 ft |
$P = 75$ | B1 ft |
$x$, $y$ and $z$ from their tableau; $P$ from their tableau, provided $P \geq 0$ | B1 |
| [2] |

## (iv)
Problem is unbounded; No limit to how big $y$ (and hence $P$) can be; Only negative in objective row is $y$ column, but all entries in this column are negative | B1 |
Any one of these, or equivalent. If described in terms of pivot choices, must be complete and convincing | B1 |
| [1] |

---
\begin{enumerate}[label=(\roman*)]
\item Represent the linear programming problem below by an initial Simplex tableau. [2]

Maximise \quad $P = 25x + 14y - 32z$,

subject to \quad $6x - 4y + 3z \leqslant 24$,
\qquad\qquad\quad $5x - 3y + 10z \leqslant 15$,

and \qquad\qquad $x \geqslant 0$, $y \geqslant 0$, $z \geqslant 0$.

\item Explain how you know that the first iteration will use a pivot from the $x$ column. Show the calculations used to find the pivot element. [3]

\item Perform one iteration of the Simplex algorithm. Show how each row was calculated and write down the values of $x$, $y$, $z$ and $P$ that result from this iteration. [7]

\item Explain why the Simplex algorithm cannot be used to find the optimal value of $P$ for this problem. [1]
\end{enumerate}

\hfill \mbox{\textit{OCR D1 2008 Q6 [13]}}