Further Maths - Simplex Algorithm
See Also
Flashcards
2022-03-30
When writing a linear programming problem in standard form for simplex, how can you convert a “minimising” into a “maximising”??
Multiply by $-1$.
What sort of constraints must be used for simplex, $\le$ or $\ge$??
\[\le\]When writing a linear programming problem in standard form for simplex, how can you convert a $f(x, y) \ge …$ into a $\le$ constraint?
Multiply by $-1$
\[-f(x, y) \le ...\]What must be true about all variables in the simplex method?
They must be $\ge$ $0$.
How would you change “maximise $P = 3x + 2y$” for the simplex algorithm?
Rearrange to
\[P - 3x - 2y = 0\] Where do you look for the pivot value?
In the column with the most negative value on the objective row.
When there are no negatives in the objective row of a simplex tableau, how do you find the optimal solution?
Pick the biggest number.
\[2x + 2y + 5z \le 60\]
How could you rewrite this using a slack variable?
What is true about all the slack variables ($r, s…$) when using the simplex method?
They are greater than or equal to $0$.
How can you spot that $P, R, S$ are basis columns?
They contain just one $1$ and the rest are $0$.
How can you read off the solution represented by a tableau?
Look for the values of the basis variables.
What is the value of the basis variable $r$ here?
The variable $x$ here is non-basic. What does that mean its value is?
$z$ has been identified as the pivot column. What are you dividing and trying to minimise in order to find the pivot row?
The final value (e.g. $60$) and the non-negative values in the pivot column (e.g. $5$).
When determining the pivot row by dividing $60/5$ or $56/2$, are you looking for the biggest number or the smallest number?
The smallest number.
Once you’ve found the pivot value (here $5$), what do you do when constructing the next tableau?
Write that row but divide all entries by the pivot value.
When constructing a new tableau in the simplex algorithm, what are you effectively doing?
Trying to turn a non-basic variable into a basic variable.
You’re creating a new tableau in the simplex algorithm in order to turn $z$ into a basic variable. This means that you’re going to need just one $1$ and the rest of the entries are going to be $0$. What do you add together to make the top row have a zero?
2022-03-31
What trick can you use to prevent errors when working out the rows of a new tableau in the simplex algorithm?
Write zeroes where you know they should exist.
2022-04-06
What do you do if there is a division by zero in the objective column in simplex?
Ignore it.
\[x + y \ge 800\]
How would you rewrite this for two-stage simplex?
\[x + y - s _ 1 + a = 800\]
What is the name of a variable like $s _ 1$?
A surplus variable.
Why must artificial variables be introduced in two-stage simplex?
So that there is a basic variable in that row.
What additional objective function do you have for two-stage simplex?
Maximise $I = -(a _ 1 + a _ 2 + a _ 3 + \ldots)$
Why is the first step of minimising the artificial variables necessary in two-stage simplex?
Minimising the artificial variables gets you to a point in the feasible region.
Why do greater than or equal to constraints like $x \ge 0$ cause an issue when using the simplex algorithm?
Because it means the origin is not in the feasible region.
When can you remove the artificial variables in two-stage simplex and move onto the next stage?
When they are all basic variables.
What is the advantage of the big M method?
You can do both stages of two-stage simplex at once.
What do you subtract from the initial objective function when using the big M method?
When can you remove the artificial variable columns when using the big M method?
When the $M$ terms are only in the artificial variable columns.
Once there are only $M$s in the artificial variable columns, what can you do?
Remove the columns.
How can you find an expression for $A$ when subtracting $MA$ from the objective function?
$A$ is equal to the sum of the artificial variables.
How can you get a solution to a linear programming problem using the big M method if there is still $M$ in the columns?
$M$ might be in all the non-basic columns, which will be $0$.
Once there’s only $M$s in non-basic columns, what do you do?
Rewrite the tableau with the artificial columns removed.
What do you do if there is a negative number during the ratio test for simplex?
Ignore it.
2022-04-30
\[3x - 2y - 2z \ge 0\]
Yuck, greater-than constraints in simplex are annoying. Why can you multiply both sides by $-1$ in this case to get a less-than constraint?
Because the origin is still in the feasible region.
\[-x-y \le -1\]
Why does this need to be changed to a greater-than constraint in order to use simplex?
Because it means that the origin is not in the feasible region.
What is the golden rule for knowing whether to use a surplus and artificial variable for a constraint in simplex?
“Does this take the origin out of the feasible region?”
2022-05-03
Why can’t simplex be used here?
Because the only negative value in the objective row is in column $z$, and all values in this column are negative and so are ignored when doing the ratio test.