You are on page 1of 51

Chapter 4

Calculus

1
 Expressions vs. functions.
 Making your own functions.
 Limits.
 Derivatives and applications.
 Taylor Polynomials.
 Integrals.
 Series.

2
 Give the file any name you want, but make sure to
save it with the .mw extension (not .mws).
 As usual, you should enter the commands from
the slides into Maple on your own.
 Also, as usual, this is NOT to be submitted for
grading.

3
 You’ve learned many calculus concepts in Math
1910 and 1920.
 You’ve seen limits, derivatives, integrals, series,
and applications of all of these concepts.
 In this chapter, we’re going to see how Maple
handles these ideas and solve a few problems
that couldn’t be done by hand in your first year
courses!

4
 Maple has many pre-defined functions, many of
which you saw in the chapter on algebra.
 Most mathematics you do are JUST on a set of
standardized functions! You usually work on some
sort of transcendental combination of them!
 You can define your own functions in two ways.
◦ An expression (you’ve already used these).
◦ A function (slightly different way, but can be more
convenient at times).

5
 To make an expression and store it in a variable,
you’ve already seen the syntax:
myExpression:=formulaInvolvingX
 Example: Make your expression 𝑥 2 :
f:=x^2
 This is good for graphing or solving an equation,
but one other thing you might need to do is
substitute values into your function, e.g., evaluate
f(5).
 What happens when you type f(5)?
 Certainly not the result you were expecting!

6
 Syntax: To substitute values into an expression,
you need to issue a somewhat bulky command:
subs(x=value, myExpression)
 Example: subs(x=5, f) is what you would use
to evaluate f(5).
 Why didn’t Maple use the standard f(5) notation?
 The answer is, it DOES, but you must define f as a
function and not an expression!

7
 Syntax: To define a function called myFunction,
of a variable called myVariable:
myFunction:=myVariable->formulaInMyVariable
 Example: f:=x->x^2
defines a function called f in the variable x and is
equal to x squared.
 Now you can use the standard notation f(5).

8
 You’ve plotted/solved expressions before. If you had
an expression called f and you wanted to plot it, you
used plot(f, x) or solve(f=number, x), etc.
 Try the same command now on your new function f:
plot(f, x) What’s happening?
 In order to plot the function, use the full notation f(x)
rather than just f. This turns the function into an
expression (internally; you won’t see any conversion
happen) and you can plot as you have done before:
plot(f(x), x)
 Similar rules apply to solving, factoring, etc.

9
 Recall shifting and scaling of functions?
 You could come up with the new formula and type
it directly in to Maple, or, use the convenient
function notation!
 Example: Take our function 𝑓 𝑥 = 𝑥 2 and plot the
new function resulting from vertically scaling the
function by a factor of -2, shifting it left by 3 units
and down by 4 units.
plot(-2*f(x+3)-4, x)

10
 As a brief note, you can also declare functions of 2 or
more independent variables (for plotting surfaces, etc).
 Syntax:
fName:=(var1, var2, …)->formula
 Example: Create and plot the function
𝜋
𝑧= 𝑥3 + sin(𝑦 2 ). Use it to evaluate 8 + sin .
2
h:=(x,y)->x^3 + sin(y^2)
plot3d(h(x, y), x=-2..2, y=-2*Pi..2*Pi)
evalf(h(2, (Pi/2)^(1/4)))

11
 Recall that a piecewise function changes its
formula depending on the value input.
 Syntax:
fName:=x->piecewise(range1, formula1,
range2, formula2, …)
 You can have as many conditions as you like.
 You can use the usual =, >, >=, <, <= symbols.
 You can graph these like any other function.

12
2
 Example: Define the function 𝑓 𝑥 = ቊ 𝑥 if 𝑥 < 1
1+𝑥 if 1 ≤ 𝑥 < 10

f:=x->piecewise(x<1, x^2, 1<=x<10,1+x)

◦ Evaluate f(-1) and f(2) to make sure things are as you


expect.
◦ Plot the function on the x-window of -5 to 15 and y-
window of 0 to 40. Note that Maple puts a vertical
line at discontinuities and at the ending point of 10
(from the definition of the function). To stop that
from happening, use the discont=true option inside
the plot command.

13
 There are easy limits (ones you can just substitute
numbers into, those that are easy to factor, etc),
and there are hard limits (squeeze theorem,
l’Hospital’s rule, etc).
 No matter what kind of limit you have, Maple has
all of these techniques programmed internally and
will automatically choose the best one to evaluate
the limit.
 You can do regular limits, or one-directional limits.

14
 Syntax:
limit(function, var=value, option)
◦ options can be left or right. If omitted, Maple
computes both and returns the value if they’re both
equal, or “undefined” otherwise.
 If you are using Maple as a word processor (as we
saw in Chapter 1) and you actually want to see the
standard limit notation, use a capital L:
Limit(function, var=value, option)
◦ This will not evaluate the limit and will merely write limit
notation.
◦ You could then evaluate it using value() or evalf().

15
 Example: Write the limit notation for the following, then
evaluate the limit: 𝑥→4
lim (2𝑥 − cos(𝜋𝑥/4)

Limit(2x-cos(Pi*x/4), x=4) #AND THEN USE:


value(%)
ln 𝑥
 Example: Evaluate lim
𝑥→∞ 𝑥

limit(ln(x)/x, x=infinity)

 (Note that Maple used l’Hospital’s rule behind the


scenes for this one!)

16
 You can also do limits on piecewise functions,
both normally and directionally.
𝜋sin(𝜋𝑥) if 𝑥 < 1
 Example: If 𝑓 𝑥 = ቊ
𝜋𝑥 + cos(𝜋𝑥) if 𝑥 ≥ 1
lim 𝑓 𝑥 , lim 𝑓(𝑥) , and lim 𝑓(𝑥)
evaluate 𝑥→1 − +
𝑥→1 𝑥→1
f:=x->piecewise(x<1, Pi*sin(Pi*x),
x>=1, Pi*x + cos(Pi*x))
limit(f(x), x=1)
limit(f(x), x=1, left)
limit(f(x), x=1, right)

17
 Maple can compute derivatives for even extremely
complicated functions.
 The standard derivative formulas for all basic functions
are all programmed into Maple, along with the product
rule, quotient rule, and chain rule.
 Maple knows when to use each rule on its own, so you
don’t have to “help” it along.
 You can get Maple to find derivatives of both
expressions and functions.
◦ The syntax is different for each type, but will produce the same
results.
◦ There’s no major advantage to using one form vs. the other.

18
 If you are working with an expression, e.g.,
f:=expression (NOT f:=x->formula), then…
 Syntax:
diff(expression, variable)
will give the first derivative of the expression.
Alternatively, you can use “normal” notation, f’
 Syntax:
diff(expression, variable$order)
will give the derivative specified by order (an integer
at least 2).
You can also use standard notation as above, for
example, f’’’ would give the third derivative.

19
 If you are using function style (f:=x->formula),
then…
 Syntax:
D(f) or f’(x)
will give the first derivative (note that it’s
f’(x)and NOT just f’ like in expression format!)
 To get higher derivatives, you need to use extra
D’s or primes. For example, D(D(D(f))) or
f’’’(x) will give the third derivative.

20
 Example: Find the derivative of
𝑓 𝑥 = 5𝑥 2 + sin(𝑥) .
f:=5x^2+sin(x)
diff(f,x)
---OR---
f:=x->5x^2+sin(x)
f’(x)
 Example: Find the third derivative of (in
𝑥 2 −1
expression format) 𝑓 𝑥 =
3𝑥 −1
diff(sqrt(x^2-1)/(3^x-1), x$3)

21
 Recall that the derivative represents the slope of
the tangent line to the graph at any value of x.
 When you substitute a number into the derivative,
the result is the slope of the line at that one point,
or, 𝑓′(𝑎).
 The equation of a tangent line to 𝑓(𝑥) at the point
𝑥 = 𝑎 is given by
𝑦 = 𝑓 𝑎 + (𝑓 ′ 𝑎 )(𝑥 − 𝑎)
 This is also called the Linear Approximation.

22
 Example: Find the slope of the tangent line to 𝑓 𝑥 =
tan(𝑥) at the point 𝑥 = 𝜋/4. There are many ways this
can be done in Maple. Here are two (one with
expression and one with function).

f:=tan(x)
fp:=f’
subs(x=Pi/4, fp)
---OR---
f:=x->tan(x)
f’(Pi/4)

23
 Example: Find the equation of the tangent line to
𝑓 𝑥 = 𝑥 3 + 2𝑥 2 − 5𝑥 − 6 at 𝑎 = −3. (Again in this
case, it might be easier to use functions)

f:=x->x^3+2x^2-5x-6
tline:=f(-3)+f’(-3)*(x-(-3))

24
 Recall that these problems involved finding the
best value for a function (global max or min) and
the value of x at which it occurs.
 When modeling a real-world situation, Maple
CANNOT come up with the function to be
optimized nor the interval of validity. YOU have to
do that!
 What Maple CAN do is find the optimal value for
you, by using the methods you learned in first year
calculus, along with graphs for visualization.

25
 Recall that the optimal value of a continuous function
occurs only at endpoints or critical points (derivative is zero
or undefined).
 Therefore, once YOU have analyzed the problem and
come up with the objective function and the constraints,
you then use Maple to do the computations to find the
optimal value.
 Once you have found the function and interval, use the
following:
◦ Syntax: maximize(expression, interval, location=true)
◦ Syntax: minimize(expression, interval, location=true)
 If an interval is something like 0 ≤ 𝑥 ≤ 4 then you need to
enter TWO intervals, i.e., 𝑥 ≥ 0, 𝑥 ≤ 4.

26
 Example: There are several pairs of positive real
numbers which add to give 100. Of all such pairs,
find the two giving the maximum possible product.

27
 Recall that all differentiable functions can be
approximated by a Taylor polynomial (any degree)
centered at a point 𝑎.
 This is merely a generalization of the linear
approximation (it’s a Taylor polynomial of degree
1).
 The higher the degree used, the more accurately it
approximates the function.
𝑓′′ 𝑎
 𝑝𝑛 𝑥 = 𝑓 𝑎 + 𝑓 ′ 𝑎 𝑥 − 𝑎 + 𝑥−𝑎 2 +
2!
𝑓3 𝑎 3 𝑓𝑛 𝑎
𝑥−𝑎 + ⋯+ 𝑥−𝑎 𝑛
3! 𝑛!

28
 You’ll recall these were not difficult, but incredibly
tedious, to do by hand.
◦ It was easy for simple functions like 𝑒 𝑥 , sin(𝑥), etc.
2
◦ But what about for a function like 𝑥𝑒 ? If you had to
−𝑥

take, say, the 5th Taylor polynomial, the product and chain
rules explode very quickly into a massive expression!
 You could use Maple’s derivative commands that
you have already learned to construct these
yourself, but Maple has a handy command that
does it all at once.

29
 Syntax:
taylor(expression, x=a, n)
Produces the Taylor polynomial OF ORDER n-1
centered at x=a.
 Note the use of the word EXPRESSION. Don’t
use function format there!

30
 Example: Compute the first, second, and third degree
𝑥−𝑥 2
Taylor Polynomials for the function 𝑓 𝑥 = 𝑥𝑒 at 𝑎 = 0.
 Then, graph 𝑓(𝑥) (black, thickness 5) and each of these
polynomials (red, blue, gold, each thickness 1) on the same
screen.
f:=x*exp(x-x^2) #expression, not function!
taylor(f,x=0, 2) #order 1
taylor(f,x=0, 3) #order 2
taylor(f,x=0, 4) #order 3
 To graph, copy and paste everything from each of these
outputs (except the 𝑂(𝑥 𝑛 ) at the end) into a plot command
with appropriate syntax that you’ve seen before.

31
 Maple has the ability to calculate both definite and
indefinite integrals.
 When possible, Maple will evaluate a definite
integral exactly. Otherwise, you can ask it to
approximate using a numerical integration rule.
 All integration methods, such as substitution,
integration by parts, etc., are done automatically.
 Syntax:
int(expression, x=a..b) → Definite
int(expression, x) → Indefinite

32
 Example: Find the antiderivative of
2 1
𝑓 𝑥 =1−𝑥 +
1−𝑥 2
int(1-x^2-1/sqrt(1-x^2), x)
 Example: Find the net area bounded by 𝑓 𝑥 =
9 − 𝑥 2 and the x-axis, between -3 and 3.
int(sqrt(9-x^2), x=-3..3)

33
 Recall that Riemann sums, and the trapezoid rule
and Simpson’s rule, are where you approximate
an integral using left/right endpoints, midpoints,
trapezoids, and parabolas, using a finite number
of sub-intervals.
 Maple has facilities by which you can draw in the
geometric figures and compute the
approximations.

34
 To show the partitioning on the graph, or to make the
computations, you need to load the Student Calculus1
package:
with(Student[Calculus1])
 Syntax:
ApproximateInt(expression, x=a..b,
method=methodType,
partition=numRectangles, options)
methodType can be left, right, midpoint,
trapezoid, simpson.
common options:
output=plot, output=animation

35
 You would not use numerical integration in Maple
for functions that have a closed form antiderivative
(such as 𝑥 2 , sin(𝑥), etc.).
◦ This is because using int() on these functions will
return an exact value, rather than an approximation.
−𝑥 2 1
 However, for other functions, like 𝑒 , , and
ln 𝑥
the like, which do not have an elementary
antiderivative, ApproximateInt() provides
good approximations.

36
14 1
 Example: Approximate ‫׬‬2 ln 𝑥 𝑑𝑥 with the
Midpoint rule, using 4 subintervals. Include the
animation plot with the outputs. Change the
frames per second to 1 and then play the
animation.
ApproximateInt(1/ln(x), x=2..14,
method=midpoint, partition=4,
output=animation)

37
Recall the absolute error bounds in the Midpoint,
Trapezoid, and Simpson’s rule were the following
maximal values:
𝑘 𝑏−𝑎 3 𝑘 𝑏−𝑎 3 𝑐 𝑏−𝑎 5
𝐸𝑀 = 2
, 𝐸𝑇 = 2
, 𝐸𝑆 =
24𝑛 12𝑛 180𝑛4
Where 𝑘 = max |𝑓 ′′ 𝑥 | , 𝑐 = max |𝑓 (4) 𝑥 |
[𝑎,𝑏] [𝑎,𝑏]
 You may recall that in math 1920, these were
“easy” to find, except for the values of k and c.
 Often you needed a graph to do this. Now we can
easily get these graphs!

38
 Example: A few slides back, we approximated
14 1
‫׬‬2 ln 𝑥 𝑑𝑥 with the midpoint rule using 4
subintervals. This gave us an approximation, but
we don’t know roughly how “good” it is. Use the
Midpoint error bound formula to get this estimate
of “goodness”.
 We know that 𝑎 = 2, 𝑏 = 14, 𝑛 = 4. We need the
𝑑2 1
value of 𝑘 = max 2 .
[2,14] 𝑑𝑥 ln 𝑥
 By hand, this is difficult. In Maple, this is much
easier!

39
 Example continued:
f:=1/ln(x)
g:=diff(f, x$2)
plot(|g|, x=1.5..12.5)
 From the graph, it appears that k, which is the farthest
distance the function travels vertically from the x-axis,
occurs at 𝑥 = 2. Thus, we can take 𝑘 = 𝑔(2):
k:=subs(x=2, g)
 Alternatively, Maple can maximize this for you:
maximize(|g|, x=2..12)
Warning: Sometimes Maple won’t be able to do it this way
because it’s too complicated Mathematically. In these
cases, Maple will either output “junk” or just seem to hang
for a long time. If this happens, just use the graph.

40
 Example continued:
So, the error is bounded by
Em:=k*(14-2)^3/(24*4^2)

41
 Recall that the length of a differentiable function
on the interval [𝑎, 𝑏] is
𝑏
2
න 1+ 𝑓′ 𝑥 𝑑𝑥
𝑎
 Due to the square root, it’s very often impossible
to find the exact length as the integral may be
impossible.
 Thus, you often have to use a numerical method
to approximate the integral.

42
 Example: If possible, find the length of the curve
ln(sec 𝑥 ) on the interval [0, 𝜋/4].
f:=ln(sec(x))
int(sqrt(1+(f’)^2), x=0..Pi/4)
 Note that the exact value was given here (actually, this
is also quite easy to do by hand!)
 Example: Repeat this example for 𝑓 𝑥 = sin(𝑥) on
[0, 𝜋].
f:=sin(x)
int(sqrt(1+(f’)^2), x=0..Pi)
 Note that this time, Maple outputs odd looking stuff!
Therefore, it’s time to approximate with a numerical
method.

43
 Example continued: Suppose we use Simpson’s
rule with n=10 to approximate this integral:
ApproximateInt(sqrt(1+(f’)^2),
x=0..Pi, partition=10, output=value)
evalf(%)

44
 Summations in Maple are similar to integration in
that Maple will try to find the exact value first. If it
can’t, it will use an approximation.
 Syntax:
sum(formula, var=a..b)
 Note: formula is anything involving the variable,
and either a or b, or both, can be infinite.
 To write the notation only, use
Sum(formula, var=a..b)
◦ This writes out the notation and does not evaluate the
summation.

45
 Maple knows formulas for computing exact values
(such as geometric series).
 It also has a few tricks up its sleeve that you likely
won’t learn until a higher-level math course (e.g.,
Fourier Series).
 Example: Evaluate the geometric series
1 3𝑛+1
σ∞
𝑛=2 2 ⋅ −
7
sum(2*(-1/7)^(3n+1), n=2..infinity)

46
 Example: When you learned p-series, you could
only get approximate values because the
techniques required to evaluate them exactly were
too advanced for first year calculus. Maple has
those techniques, so, evaluate the p-series
∞ 1
σ𝑛=1 2
𝑛
sum(1/n^2, n=1..infinity)

47
 There are, however, some series that Maple can’t
do. It won’t even give you an approximate answer.
In this case, you have to approximate it yourself.
∞ 1
 Example: Evaluate σ𝑛=2 2 if possible. If not,
𝑛 ln 𝑛
use a finite number of terms to approximate the
series to within 0.001 of its exact value using trial
and error (keep increasing n until the value of the
series changes by less than 0.001).

48
 Example continued:
sum(1/(n^2*ln(n)), n=2..infinity)
 Maple refuses to evaluate it, even with evalf()!
 Therefore, approximate. Start with, say, a range of 2 to 10
and then go to 11 and compare:
sum(1/(n^2*ln(n)), n=2..10)
sum(1/(n^2*ln(n)), n=2..11)
 The difference is more than 0.001, so go to 12, and so on,
until the difference is less than 0.001. It’s quickest to go up
in increments of more than 1 (say 10), and then check the
next one for the difference.
sum(1/(n^2*ln(n)), n=2..20)
sum(1/(n^2*ln(n)), n=2..21)

49
 Example continued: The difference here is less
than 0.001, so use the approximation from 2 to 21
as the approximate value.

50
 Calculus has a wide range of applications, and
we’ve seen how to get Maple to do the grunt work
for us. This again only scratches the surface.
 Next up: Programming!
 Now, don’t forget to finish the Chapter 4
Assignment and submit to Moodle by the indicated
deadline!

51

You might also like