2.1: Derivatives
( \newcommand{\kernel}{\mathrm{null}\,}\)
We are now ready to explain what is meant by the slope of a curve or the velocity of a moving point. Consider a real function f and a real number a in the domain of f. When x has value a, f(x) has value f(a). Now suppose the value of x is changed from a to a hyperreal number a+Δx which is infinitely close to but not equal to a. Then the new value of f(x) will be f(a+Δx). In this process the value of x will be changed by a nonzero infinitesimal amount x, while the value of f(x) will be changed by the amount f(a+Δx)−f(a).
The ratio of the change in the value of f(x) to the change in the value of x is f(a+Δx)−f(a)Δx.
This ratio is used in the definition of the slope of f which we now give.
S is said to be the slope of f at a if S=st(f(a+Δx)−f(a)Δx)for every nonzero infinitesimal Δx.
The slope, when it exists, is infinitely close to the ratio of the change in f(x) to an infinitely small change in x. Given a curve y=f(x), the slope of f at a is also
called the slope of the curve y=f(x) at x=a. Figure 2.1.1 shows a nonzero infinitesimal Δx and a hyperreal straight line through the two points on the curve at a and a+Δx. The quantity f(a+Δx)−f(a)Δx is the slope of this line, and its standard part is the slope of the curve.

The slope of f at a does not always exist. Here is a list of all the possibilities.
- The slope of f at a exists if the ratio f(a+Δx)−f(a)Δx is finite and has the same standard part for all infinitesimal Δx≠0. It has the value S=st(f(a+Δx)−f(a)Δx).
- The slope of f at a can fail to exist in any of four ways:
- f(a) is undefined.
- f(a+Δx) is undefined for some infinitesimal Δx≠0.
- The term f(a+Δx)−f(a)Δx is infinite for some infinitesimal Δx≠0.
- The term f(a+Δx)−f(a)Δx has different standard parts for different infinitesimals Δx≠0.
We can consider the slope of f at any point x, which gives us a new function of x.
Let f be a real function of one variable. The derivative of f is the new function f′ whose value at x is the slope of f at x. In symbols, f′(x)=st(f(x+Δx)−f(x)Δx)whenever the slope exists.
The derivative f′(x) is undefined if the slope of f does not exist at x.
For a given point a, the slope of f at a and the derivative of f at a are the same thing. We usually use the word "slope" to emphasize the geometric picture and
"derivative" to emphasize the fact that f′ is a function.
The process of finding the derivative of f is called differentiation. We say that f is differentiable at a if f′(a) is defined; i.e., the slope of f at a exists.
Independent and dependent variables are useful in the study of derivatives. Let us briefly review what they are. A system of formulas is a finite set of equations and inequalities. If we are given a system of formulas which has the same graph as a simple equation y=f(x), we say that y is a function of x, or that y depends on x, and we call x the independent variable and y the dependent variable.
When y=f(x), we introduce a new independent variable Δx and a new dependent variable Δy, with the equation Δy=f(x+Δx)−f(x).
This equation determines Δy as a real function of the two variables x and Δx, when x and Δx vary over the real numbers. We shall usually want to use the Equation (2.1.1) for Δy when x is a real number and Δx is a nonzero infinitesimal. The Transfer Principle implies that Equation (2.1.1) also determines Δy as a hyperreal function of two variables when x and Δx are allowed to vary over the hyperreal numbers.
Δy is called the increment of y. Geometrically, the increment Δy is the change in y along the curve corresponding to the change Δx in x. The symbol y′ is sometimes used for the derivative, y′=f′(x). Thus the hyperreal equation f′(x)=st(f(x+Δx)−f(x)Δx)now takes the short form y′=st(ΔyΔx)
The infinitesimal Δx may be either positive or negative, but not zero. The various possibilities are illustrated in Figure 2.1.2 using an infinitesimal microscope. The signs of Δx and Δy are indicated in the captions.

Our rules for standard parts can be used in many cases to find the derivative of a function. There are two parts to the problem of finding the derivative f′ of a function f:
- Find the domain of f′.
- Find the value of f′(x) when it is defined.
Find the derivative of the function f(x)=x3.
Solution
In this and the following examples we let x vary over the real numbers and Δx vary over the nonzero infinitesimals. Let us introduce the new variable y with the equation y=x3. We first find Δy/Δx.
y=x3,y+Δy=(x+Δx)3,Δy=(x+Δx)3−x3,ΔyΔx=(x+Δx)3−x3Δx.
Next we simplify the expression for Δy/Δx.
ΔyΔx=(x3+3x2Δx+3x(Δx)2+(Δx)3)−x3Δx=3x2Δx+3x(Δx)2+(Δx)3Δx=3x2+3xΔx+(Δx)2.
Then we take the standard part, st(ΔyΔx)=st(3x2+3xΔx+(Δx)2)=st(3x2)+st(3xΔx)+st((Δx)2)=3x2+0+0.
Therefore, f′(x)=st(ΔyΔx)=3x2.
We have shown that the derivative of the function f(x)=x3is the function f′(x)=3x2 with the whole real line as domain. f(x) and f′(x) are shown in Figure 2.1.3.

Find f′(x) given f(x)=√x.
Solution
Case 1: x<0
Since √x is not defined, f′(x) does not exist.
Case 2: x=0
When Δx is a negative infinitesimal, the term √x+Δx−√xΔx=√0+Δx−√0Δx is not defined because √Δx is undefined. When Δx is a positive infinitesimal, the term √x+Δx−√xΔx=√ΔxΔx=1√Δxis defined but its value is infinite. Thus for two reasons, f′(x) does not exist.
Case 3: x>0
Let y=√x. Then y+Δy=√x+Δx,Δy=√x+Δx−√x,ΔyΔx=√x+Δx−√xΔx.
We then make the computation ΔyΔx=(√x+Δx−√x)Δx⋅(√x+Δx+√x)(√x+Δx+√x)=(x+Δx)−xΔx(√x+Δx+Δx)=ΔxΔx(√x+Δx+√x)=1√x+Δx+√x.
Taking standard parts, st(ΔyΔx)=st(1√x+Δx+√x)=1st(√x+Δx+√x)=1st(√x+Δx)+st(√x)=1√x+√x=12√x.
Therefore, when x>0, f′(x)=12√x.
So, the derivative of f(x)=√x is the function f′(x)=12√x, and the set of all x>0 is its domain (see Figure 2.1.4.

Find the derivative of f(x)=1/x.
Solution
Case 1: x=0
Then 1/x is undefined so f′(x) is undefined.
Case 2: x≠0
y=1/x,y+Δy=1x+Δx,Δy=1x+Δx−1x,ΔyΔx=1/(x+Δx)−1/xΔx.
Simplifying, 1/(x+Δx)−1/xΔx=x−(x+Δx)x(x+Δx)Δx=−Δxx(x+Δx)Δx=−1x+Δx
Taking the standard part, st((ΔyΔx)=st(−1x(x+Δx))=−1st(x(x+Δx))=−1st(x)st(x+Δx)=−1x⋅x=−1x2.
Thus, f′(x)=−1x2.
The derivative of the function f(x)=1/x is the function f′(x)=−1/x2 whose domain is the set of all x≠0. Both functions are graphed in Figure 2.1.5.

Find the derivative of f(x)=|x|.
Solution
Case 1 x>0.
In this case |x|=x, and we have y=x,y+Δy=x+Δx,Δy=Δx,ΔyΔx=1,f′(x)=1.
Case 2: x<0.
Now |x|=−x, and y=−x,y+Δy=−(x+Δx),Δy=−(x+Δx)−(−Δx)=−Δx,ΔyΔx=−ΔxΔx=−1,f′(x)=−1.
Case 3: x=0.
Then y=0,y+Δy=|0+Δx|=|Δx|,Δy=|Δx|,ΔyΔx=|Δx|Δx={1if Δx>0,−1if Δx<0.
The standard part of Δy/Δx is then 1 for some values of Δx and −1 for others. Therefore f′(x) does not exist when x=0.
In summary, f′(x)={1if x>0,−1if x<0,undefinedif x=0
Figure 2.1.6 shows f(x) and f′(x).

The derivative has a variety of applications to the physical, life, and social sciences. It may come up in one of the following contexts.
Velocity: If an object moves according to the equation s=f(t) where t is time and s is distance, the derivative v=f′(t) is called the velocity of the object at
time t.
Growth rates: A population y (of people, bacteria, molecules, etc.) grows according to the equation y=f(t) where t is time. Then the derivative y′=f′(t) is
the rate of growth of the population y at time t.
Marginal values (economics): Suppose the total cost (or profit, etc.) of producing x items is y=f(x) dollars. Then the cost of making one additional item is approximately the derivative y′=f′(x) because y′ is the change in y per unit change in x. This derivative is called the marginal cost.
A ball thrown upward with initial velocity b ft per sec will be at a height y=bt−16t2feet after t seconds. Find the velocity at time t. Let t be real and Δt≠0, infinitesimal.
Solution
y+Δy=b(t+Δt)−16(t+Δt)2,Δy=[b(t+Δt)−16(t+Δt)2]−[bt−16t2],ΔyΔt=[b(t+Δt)−16(t+Δt)2]−[bt−16t2]Δt=bΔt−32tΔt−16(Δt)2Δt=b−32t−16Δt.
At time t sec, v=y′=b−32 ft/sec.
Both functions are graphed in Figure 2.1.7.

Suppose a bacterial culture grows in such a way that at time t there are t3 bacteria. Find the rate of growth at time t=1000 sec.
Solution
y=t3,y′=3t2 by Example 2.1.1.
At t=1000 sec, y′=3,000,000 bacteria/sec.
Suppose the cost of making x needles is √x dollars. What is the marginal cost after 10,000 needles have been made?
Solution
y=√x,y′=12√x by Example 2.1.2.
At x=10,000, y′=12√10,000=1200 dollars per needle.
Thus the marginal cost is one half of a cent per needle.
Problems for Section 2.1
Find the derivative of the given function in Problems 1-21.
1. | f(x)=x2 | 2. | f(t)=t2+3 |
3. | f(x)=1−2x2 | 4. | f(x)=3x2 |
5. | f(t)=4t | 6. | f(x)=2−5x |
7. | f(t)=4t3 | 8. | f(t)=−t3 |
9. | f(u)=5√u | 10. | f(u)=√u+2 |
11. | g(x)=x√x | 12. | g(x)=1/√x |
13. | g(t)=t−2 | 14. | g(t)=t−3 |
15. | f(y)=3y−1+4y | 16. | f(y)=2y3+4y2 |
17. | f(x)=ax+b | 18. | f(x)=ax2 |
19. | f(x)=√ax+b | 20. | f(x)=1/(x+2) |
21. | f(x)=1/(3−2x) | ||
22. | Find the derivative of f(x)=2x2 at the point x=3. | ||
23. | Find the slope of the curve f(x)=√x−1 at the point x=5. | ||
24. | An object moves according to the equation y=1/(t+2), t≥0. Find the velocity as a function of t. | ||
25. | A particle moves according to the equation y=t4. Find the velocity as a function of t. | ||
26. | Suppose the population of a town grows according to the equation y=100t+t2. Find the rate of growth at time t=100 years. | ||
27. | Suppose a company makes a total profit of 1000x−x2 dollars on x items. Find the marginal profit in dollars per item when x=200, x=500, and x=1000. | ||
28. | Find the derivative of the function f(x)=|x+1|. | ||
29. | Find the derivative of the function f(x)=|x3|. | ||
30. | Find the slope of the parabola y=ax2+bx+c where a, b, c are constants. |