Local maximum and minimum points
In a minute
At a or inside the of a smooth curve, the is zero, so you find these points by solving . On Paper 2 your GDC can draw from and solve the equation for you. A local maximum need not be the greatest value on the : check the end points too.
You'll be able to
- Solve to find the -values at which a curve has a horizontal .
- Use your GDC to generate from and solve , writing down the equation you solve.
- Find the coordinates of local maximum and minimum points, decide which is which, and say what they mean in context.
- Find the greatest and least values of a model on its domain, checking the end points as well as the local maximum and minimum points.
Before you start3 quick questions on what this builds on
Revise first: 2.1 Straight-line equations, 2.2 Function concepts, domain and range, 2.4 Key features of graphs, 5.1 Limits and the derivative as a rate of change, 5.2 Increasing and decreasing functions, 5.3 Differentiation of integer-power polynomials and 5.4 Tangents and normals
Where the gradient is zero
The graph below shows a fish population. It falls to a low point, then climbs to a high point. At each turning point the curve is flat for an instant: the there is horizontal.
A horizontal tangent has gradient 0. So at a local maximum or local minimum of a smooth curve, . In 5.2 you called points like this .
The reverse is not always true. only says the tangent is flat. You saw in 5.2 that is flat at the origin but rises on both sides of it.
1. Find and solve . The solutions are the -values where the gradient is zero. Keep only those in the domain.
2. Substitute each solution into , not , to find the -coordinate.
3. Decide what each point is, from the graph or from the sign of just either side. Test values closer to the point than any other solution, with no break in the graph in between.
Try it first, then reveal one step at a time.
- Differentiate: .
- Set and take out the common factor 6: , so . So or .
- Find the -coordinates from : and .
- Test the sign of on each side: , and .
- So is a local maximum, where the sign goes from to . And is a local minimum, where it goes from to .
Try it first, then reveal one step at a time.
- (a) .
- (b) Solve . Divide both sides by : , so and or . Both are in the domain.
- Find the -values: and .
- Check the types with the sign of : , and . So is the local minimum and is the local maximum, as the graph shows.
- (c) The population falls for the first 2 years, to a low of 4800 fish. It then recovers to 8000 fish after 6 years, its starting level, before it starts to fall again.
Solving $f'(x) = 0$ with technology
On Paper 2 a model can be hard to differentiate, or hard to solve by hand. Your GDC does that work. You still write down the equation you are solving, then give the GDC's values.
There are two routes. If is a sum of integer powers of , differentiate it by hand as in 5.3. Then let the GDC solve : with its polynomial solver if is a polynomial, or by graphing if it has a negative power, such as .
If you can't differentiate , for example because it contains , let the GDC generate . It can't write as a formula, but its numerical derivative tool draws the graph of from . The zeros of that graph are the solutions of .
Try it first, then reveal one step at a time.
- (a) .
- (b) Write down the equation: . The GDC's polynomial solver gives or , so or .
- (c) Test the first solution: and . The sign goes from to , so gives the local maximum.
- Use the full value in : . Store the GDC's value instead of retyping 2.16.
- So the local maximum point is . The top of the hill is about 2.16 km along the trail, 315 m above sea level.
You can check part (c) another way. The GDC's maximum tool on finds the same point, because the local maximum is where . Set a window that shows the hill first: the heights are near 300, far above the standard window.
Try it first, then reveal one step at a time.
- (a) You can't differentiate by hand at SL, so let the GDC generate . Enter as the first function and its numerical derivative as the second.
- Write down what you solve: , from the graph of for .
- The graph of crosses the -axis twice. The zero tool gives and , so or .
- Check: the GDC's maximum and minimum tools on give turning points at the same values of .
- (b) The graph of goes from above the axis to below it at , so this is a local maximum. About 6.45 months after launch, the number of subscribers stops rising and starts to fall for a while.
Local is not always the greatest or least
A local maximum is only the highest point near it. On a restricted domain, the graph may climb higher somewhere else, often at an end point. The same goes for a local minimum.
So list the candidates: the local maximum and minimum points, and both end points. Work out at each and compare the values.
Try it first, then reveal one step at a time.
- (a) , so at and .
- and . The graph rises to and then falls, so the local maximum point is .
- (b) Work out the end points too: and .
- Compare all four values: , , and . The greatest is , at .
- Comment: the local maximum at is only the best output among those near 300 items. Making 1000 items a week gives the greatest profit, $110 000.
Try it first, then reveal one step at a time.
- A local minimum has higher points right next to it, so it can't be the greatest value. The candidates are the local maximum and the two end points.
- Local maximum: . Use the stored GDC value of , not 6.45.
- End points: and .
- The greatest is : about 102 000 subscribers, at the end of the 30 months.
- Comment: the peak at about 6.45 months is only a local maximum. After a dip, the subscriber numbers climb higher by month 30. The model says nothing about what happens after 30 months.
To find on the GDC, set a window that includes first: the TI-84's value tool only accepts an inside the window.
Common mistakes
- Substituting the -value into to find the -coordinate. That always gives 0; use .
- Giving the local maximum as the greatest value of a model on a restricted domain. Work out the value at each end point and compare.
- Writing only a GDC answer. Write down the equation with your derivative in it, so the method mark can be awarded.
- Rounding before substituting it into . Keep the GDC's full value and round only the final answer to 3 s.f.
- Answering with when the question asks for the maximum value. "Find the maximum value" wants there; "find where" wants .
- Keeping a solution of that lies outside the domain, such as a negative time. Reject it and say why.
Exam tips
- "Find the coordinates" of a local maximum or minimum: give both, , with the -value worked out from and given exactly or to 3 s.f.
- "Use your answer to (a) to find" or "hence": set the derivative you found equal to zero and show that line. Using only the GDC's maximum tool on may not earn the method marks.
- "Justify" that a point is a local maximum: give the sign of just either side, such as and either side of .
Key takeaways
Practise this
39 past-paper questions on Applications of differentiation in the AI HL bank, each with a mark scheme and a worked explanation.