Mc1–Mc4 - Derivatives as gradients and rates of change

Mc1–Mc4 - Derivatives as gradients and rates of change

Turn gradients between pairs of points into the gradient at one point. Mc1–Mc3 connect that tangent direction to the derivative, its notation and rates with units; Mc4 develops the gradient graph from the shape of a curve, including cases where flattening does not reverse its direction.

From a chord to a tangent

Can you measure a curve's slope at one point without pretending the curve is straight? On y=x2y=x^2, the chord from (1,1)(1,1) to (2,4)(2,4) has gradient 33, but the chord to (1.1,1.21)(1.1,1.21) has gradient 2.12.1. Bringing the second point towards the first reveals a local direction.

You already know

For two distinct horizontal coordinates, a straight line's gradient is change in height divided by change in horizontal coordinate. A positive gradient rises to the right, a negative one falls, and a horizontal line has gradient zero.

Use a moving point on the curve, not a point chosen freely in the plane. Its height must change with its horizontal coordinate.

Chords approaching the tangent at a fixed parabola point

Chord and tangent

A chord joins two distinct curve points. A tangent, at a smooth point where a finite limiting chord gradient exists, is the line through the point with that gradient. It describes the local direction; it need not meet the curve only once. For example, the tangent to y=x3y=x^3 at (1,1)(1,1) meets the curve again at (−2,−8)(-2,-8).

There is another useful surprise: lifting the entire graph by 11 changes every height but none of its gradients. The two heights in a chord calculation both gain 11, so their difference is unchanged. That observation will help separate height from slope.

Tangent as a chord-gradient limit

For A=(a,f(a))A=(a,\mathrm f(a)), when the same finite limit exists from both sides,

f′(a)=lim⁡h→0f(a+h)−f(a)h.\mathrm f'(a)=\lim_{h\to0}\frac{\mathrm f(a+h)-\mathrm f(a)}{h}.

Simplify with h≠0h\ne0, then take the limit. The first-principles limit definition is supplied.

In the formula sheet.

A table suggests a value, but exact simplification explains why that value is approached.

Worked example 1: nearby points on a parabola

Question. Determine the tangent gradient to y=x2y=x^2 at (1,1)(1,1) by considering the point (1+h,(1+h)2)(1+h,(1+h)^2).

Plan. Express the chord gradient in terms of hh, then let the increment tend to zero.

m=(1+h)2−1hmatching coordinate changes=2h+h2h=2+hexpand; cancel for h≠0lim⁡h→0m=2the same from both sides\begin{aligned} m&=\frac{(1+h)^2-1}{h} && \text{matching coordinate changes}\\ &=\frac{2h+h^2}{h}=2+h && \text{expand; cancel for }h\ne0\\ \lim_{h\to0}m&=2 && \text{the same from both sides} \end{aligned}

Why this step? Taking the limit of the simplified expression avoids the undefined quotient obtained by substituting zero too early.

Answer. The tangent gradient is 22.

Check. The chord gradients for h=1,0.1,0.01h=1,0.1,0.01 are 3,2.1,2.013,2.1,2.01, and for h=−0.1,−0.01h=-0.1,-0.01 are 1.9,1.991.9,1.99.

Now change only the graph's vertical position. The chord calculation tests whether a higher graph must be steeper.

Worked example 2: the same slope at a different height

Question. Determine the tangent gradient to y=x2+1y=x^2+1 at (1,2)(1,2) using a nearby point.

Plan. Use (1+h,(1+h)2+1)(1+h,(1+h)^2+1) and subtract the whole fixed height.

m=(1+h)2+1−2hboth points lie on the lifted curve=2h+h2h=2+hthe height shift cancelslim⁡h→0m=2take the limit\begin{aligned} m&=\frac{(1+h)^2+1-2}{h} && \text{both points lie on the lifted curve}\\ &=\frac{2h+h^2}{h}=2+h && \text{the height shift cancels}\\ \lim_{h\to0}m&=2 && \text{take the limit} \end{aligned}

Why this step? A translation adds equally to both heights, leaving their difference unchanged.

Answer. The tangent gradient is still 22, although the point's height is now 22.

Check. At h=0.1h=0.1, the chord gradient is (2.21−2)/0.1=2.1(2.21-2)/0.1=2.1, as before.

The limit always concerns nonzero intervals. We are not replacing the chord by two identical points and then calculating a gradient.

Watch out

Wrong: substitute h=0h=0 into the unsimplified difference quotient. Right: simplify while h≠0h\ne0, then take the limit. A zero denominator is not a very small denominator.

Keep the lifted graph and change its fixed input.

The derivative at a point and as a function

A derivative packages all the available tangent gradients into one function. At a general input, the chord gradient for y=x2y=x^2 is 2x+h2x+h; its limit is 2x2x. The lifted curve y=x2+1y=x^2+1 has the same gradient function.

Derivative

For y=f(x)y=\mathrm f(x), the derivative is written f′(x)\mathrm f'(x) or dy/dx\mathrm dy/\mathrm dx. It gives the tangent gradient at every input where that finite derivative exists. The value f(a)\mathrm f(a) is a height; f′(a)\mathrm f'(a) is a gradient. For example, f(3)=9\mathrm f(3)=9 and f′(3)=6\mathrm f'(3)=6 for f(x)=x2\mathrm f(x)=x^2.

A formula alone does not remove domain restrictions. Only evaluate a derivative where the function and derivative exist. At a domain endpoint, a one-sided rate may exist without a two-sided derivative.

Worked example 3: interpreting two outputs

Question. Given g(x)=x2+1\mathrm g(x)=x^2+1 and g′(x)=2x\mathrm g'(x)=2x, interpret g(2)\mathrm g(2) and g′(2)\mathrm g'(2) geometrically.

Plan. Evaluate the height and gradient separately.

g(2)=22+1=5curve heightg′(2)=2(2)=4tangent gradient\begin{aligned} \mathrm g(2)&=2^2+1=5 && \text{curve height}\\ \mathrm g'(2)&=2(2)=4 && \text{tangent gradient} \end{aligned}

Why this step? The prime changes what the output represents, not the input being used.

Answer. The graph passes through (2,5)(2,5), with tangent gradient 44, rising locally as xx increases.

Check. Its nearby chord gradient is 4+h4+h, tending to 44.

Change only the sign of the input. The height remains positive while the direction reverses.

Worked example 4: a negative gradient above the axis

Question. Interpret g(−2)\mathrm g(-2) and g′(−2)\mathrm g'(-2) for the same function.

Plan. Substitute carefully, keeping brackets around the negative input.

g(−2)=(−2)2+1=5height stays positiveg′(−2)=2(−2)=−4gradient changes sign\begin{aligned} \mathrm g(-2)&=(-2)^2+1=5 && \text{height stays positive}\\ \mathrm g'(-2)&=2(-2)=-4 && \text{gradient changes sign} \end{aligned}

Why this step? The sign of the gradient describes direction; the sign of the height describes position relative to the axis.

Answer. The graph passes through (−2,5)(-2,5) and is locally falling to the right with tangent gradient −4-4.

Check. At x=−1.9x=-1.9, its height is 4.614.61, lower than 55.

You may also meet a derivative value in a table without knowing a formula. It has exactly the same meaning.

Watch out

Wrong: read dy/dx\mathrm dy/\mathrm dx as y/xy/x. Right: it denotes a limit of ratios of changes. On y=x2y=x^2, at x=3x=3, the coordinate ratio is 33, while the derivative is 66.

Interpret the following information as two separate graph facts.

Rates of change and their units

Suppose a cooling model gives T(t)=30−t2T(t)=30-t^2, where TT is in degrees Celsius and tt is in minutes, for 0≤t≤40\le t\le4. Between two times, the temperature difference gives an average rate. At a single interior time, the derivative gives the instantaneous cooling rate.

Continue this lesson.

You’re previewing NovaLearn without an account. Create a free account to read the rest.

Create free account