Sie sind auf Seite 1von 23

1

Logarithms and Exponentials

Paul Damien

Contents

History of Logs

Meaning of Log

Exponential and Logarithmic Functions

Applications to Interest Rates

Applications to Mortgages

Applications to Seismology

Applications to Growth and Decay

Applications to Regression

Sound, Creation, and Beethoven’s 9th Symphony


2

History of Logarithms

Hundreds of years ago before calculators were invented, people got tired of multiplying
and dividing large numbers. And so they invented Logarithms to do the calculations.
They created a table called Log Tables that could approximately multiply and divide
large numbers. But as time went on people realized that logarithms were as profound as
life itself: more on this point later. The word “log” was used in a connotation similar to
one found in keeping a “log of your work.”

Logarithms were invented independently by John Napier, a Scotsman, and by Joost


Burgi, a Swiss. The logarithms which they invented differed from each other and from
the common and natural logarithms now in use. Napier's logarithms were published in
1614; Burgi's logarithms were published in 1620. The objective of both men was to
simplify mathematical calculations. Napier's approach was algebraic and Burgi's
approach was geometric. Neither man had a concept of a logarithmic base. Napier
defined logarithms as a ratio of two distances in a geometric form, as opposed to the
current definition of logarithms as exponents. The possibility of defining logarithms as
exponents was recognized by John Wallis in 1685 and by Johann Bernoulli in 1694.

The invention of the common system of logarithms is due to the combined effort of
Napier and Henry Biggs in 1624. Natural logarithms first arose as more or less accidental
variations of Napier's original logarithms. Their real significance was not recognized
until later. The earliest natural logarithms occur in 1618.

Logarithms

Multiplication is a shortcut for addition. Recall that 3 times 5 means 5 + 5+ 5. Exponents


are a shortcut for multiplication. Recall that means . Logarithm is a shortcut
for exponents.

Powers of ten are the building blocks for logarithms. So before we reveal what
logarithms actually are, let's just explore some powers of 10:

Number 10x Power


1000000 106 6
3
1000 10 3
2
100 10 2
1
10 10 1
0
1 10 0
-1
0.1 10 -1
0.01 10-2 -2
3

0.001 10-3 -3
-6
0.000001 10 -6

You should be able to see that there is a simple relationship between the power of ten and
the number of zeros in the number. If the number is greater than 1, then the power of ten
is just the number of zeros in the number: 1000 = 103. If the number is less than 1, then
the power of ten is just minus the number of digits after the decimal point: 0.0001 = 10-4.

We are at last ready to reveal what logarithms are.

The logarithm of a number is simply


the size of the power of 10 that equals the number.

Let's take a look at that last table again with one important change:

x 10x Logarithm of x
1000000 106 6
3
1000 10 3
100 102 2
1
10 10 1
0
1 10 0
-1
0.1 10 -1
-2
0.01 10 -2
-3
0.001 10 -3
-6
0.000001 10 -6

Since 100 is 102, then the logarithm of 100 is just 2. Since 0.001 is 10-3, then the
logarithm of 0.001 is just -3.

There, that wasn't so bad was it!

Here is an equation that might help.

Definition of Logarithms
'y' is the logarithm of 'x' if 10y = x.
Remember this!

So 3 is the logarithm of 1000, because 103 is 1000.

We have established that logarithms are just another name for a power of ten. We say that
100 is '10 to the power 2', but we can equally say the logarithm of 100 is 2. Similarly
4

when we make a statement such as 'the logarithm of 2 is 0.3' what we mean is that 100.3 is
equal to 2. The only difficult idea is fractional powers of ten. But then 100 is 1, and 101 is
10, so it should seem pretty unremarkable that 100.3 is a number like 2.

What are the logarithms for other small numbers?

In particular we know the logarithm of 1, 2 and 10 (they're 0.0, 0.3 and 1.0), so how can
we find the logarithm of, say, 4?

We can find the answer very easily by observing the parallels between these two
equations:

2x2=?
10 x 100.3 = 10?
0.3

The answer to the first is 4. The answer to the second is, I hope you see, 100.6. Yes, we
just use the rule for adding powers as we did at the start of the tutorial. But look! 100.3 is 2
! The two equations are the same equations! The left hand sides are the same, so the
right hand sides are the same, so 4 is equal to 100.6, so the logarithm of 4 is just 0.6.

Let's try one more, a little more tricky. What is the logarithm of 5?

Again, let us draw a parallel between two equations:

2 x ? = 10
10 x 10? = 101
0.3

The ? in the first must stand for 5. The ? in the second must stand for 0.7 (since we have a
powers of 10 multiplication and the powers must add: 0.3+0.7=1.0). Once more the two
equations are really the same, 2 is 100.3, 10 is 101, therefore 5 is 100.7, therefore the
logarithm of 5 is 0.7.

Multiplication

Our playing at finding logarithms has a purpose. It has led us to the idea that we can
relate the multiplication of numbers to the addition of logarithms. Let me show you why.

Here’s 3.15 x 3.74:

1. Take the logarithm of 3.15 (=0.4983, so my calculator says)


2. Take the logarithm of 3.74 (=0.5729)
3. Add the logarithms together (=1.0712)
4. Calculate what 10 is raised to that power (101.0712 = 11.781)
5

5. That's the product of 3.15 and 3.74 (=11.781)

We were able to do the multiplication by using 3 simple operations: finding the logarithm
of a number, adding logarithms, finding the value of a power of 10. This last operation
(step 4) is sometimes given the horrible name of the 'antilogarithm' calculation.

So here's the general rule for doing a multiplication using logarithms:

The product of numbers A and B


is just the antilogarithm
of the sum of the
logarithm of A
and the logarithm of B:

A x B = 10(log(A)+log(B))

This looks quite unmemorable, so just remember the steps in the calculation above. To
multiply two numbers together using logarithms, first take the logarithm of each of the
numbers, then add these logarithms together, then take the antilogarithm (that is the
power of ten) of the sum, the result is the product of the two numbers.

Example: Suppose you were to multiply 1000 x 100. Obviously this is easy to do in your
head! But look at HOW you do it. You have THREE 0s and TWO 0s and you ADD
them, get FIVE 0s, and so the answer is 100,000.

But the cunning trick of logarithms is that Napier found out you could have HALF of a 0,
or TWO-THIRDS of 0s, etc, and add them up like that and get sensible answers! That’s
one of the reasons they’re lovely!

Half a 0 is the SQUARE ROOT of 10. Why? Because adding two of them makes ONE
0.

For any number you start with the question: "HOW MANY 0s?" The answer will usually
be some funny decimal.

Example: 125 times 9972 =1246500. Let’s see how we get an answer close to this using
logs.

125: how many 0s has that got? If it were 100 it would be 2 precisely.

Log table reveals LOG(125) = 2.09691

9972: how many 0s? Nearly four. (If it was 10,000 it would have exactly four zeros)

Log table reveals LOG(9972) = 3.99878


6

Add them together. 2.09691 + 3.99878 = 6.09569

The answer is going to be six and-a-bit more 0s. That is, roughly more than a million.

Antilog(6.09569) = 10 raised to the power of 6.09569 = 1246493

So that's the answer to five decimal places: 1246500 + or - a bit.

Note the word antilog simply means you’re taking the number and raising it to the power
of 10.

Division

Let us say we want to divide 15 by 5, and let us say we want to do it the hard way!

1. Take the logarithm of 15: log(15)=1.18


2. Take the logarithm of 5: log(5)=0.7
3. Subtract the logarithm of 5 from the logarithm of 15: log(15)-
log(5) = 1.18 - 0.7 = 0.48
4. Undo the logarithm by taking the power of ten: 100.48 = 3

If you find this hard, try to remember that logarithms are just a short hand for powers of
ten. When we are dividing 15 by 5 this way, really we are just doing this problem with
powers of ten:

15 ÷ 5 = 101.18 ÷ 100.7 = 100.48 = 3

Here's a summary:

The division of number A by B


is just the antilogarithm of
(the logarithm of A minus the logarithm
of B):

A ÷ B = 10(log(A)-log(B))

Functions

The logs we have worked with so far were powers of 10. They are called Log to the Base
10. But instead of working with powers of 10 what if we worked with powers of the
7

number e=2.718281828…. “e” is called the exponential number. The inverse of this
base is called the Natural Logarithm.

Before we review exponential and logarithmic functions, let's review the definition of a
function and the graph of a function. A function is just a rule. The rule links one number
to a second number in an orderly and specific manner. All the points on the graph of a
function are made up of two parts: (a number, and the function value at that number). For
example, the number of hours worked in a week could be the first number, and the salary
for the week could be the function value. If an hourly salary is $7.00, then the rule would
be 7 times the number of hours worked.

You could identify a point on the graph of a function as (x,y) or (x, f(x)). You may have
only one function value for each x number.

If the points (2, 3), (4, 5), (10, 11), and (25, 26) are located on the graph of a function,
you could easily figure out a corresponding rule. To get the function value, you just add 1
to the first number. The rule is f(x) = x + 1.
8

The points (3, 8) and (3, 18) could not be points on the graph of a function because there
are two different function values for the same x value

The inverse of the exponential function -- The Natural Logarithm

The graph of the function

y = ex

clearly shows that it is a one to one function, hence an inverse exists. We call this inverse
the natural logarithm and write it as

y = ln x

Below is the graph of

y = ex

and

y = ln x

By definition, they are reflections of each other across the line y = x.


9

Inverse Properties of Logs

Since logs and exponents cancel each other we have:

eln x = x
and

ln ex = x
Example

eln 3 = 3 and ln(e5) = 5

Three Properties of Logs

Property 1: ln (uv) = ln u + ln v (The Product to Sum Rule)

Property 2: ln (u/v) = ln u - ln v (The Quotient to Difference Rule)

Property 3: ln ur = r ln u (The Power Rule)


10

Properties of Exponentials

• ex+y = (ex)(ey)
• exy = (ex)y
• e0 = 1
• e-x = 1/(ex)
11

Applications to Interest Rates

The Compound Interest Equation

P = C (1 + r/n) nt

where
P = future value
C = initial deposit
r = interest rate (expressed as a fraction: e.g. 0.06)
n = # of times per year interest is compounded
t = number of years invested

Simplified Compound Interest Equation

When interest is only compounded once per year (n=1), the equation simplifies to:

P = C (1 + r) t

Continuous Compound Interest

When interest is compounded continually (i.e. n --> ), the compound interest equation
takes the form:

P = C e rt , where “e” is the exponential number.

Example: A $1,000 investment is made in a trust fund at an annual percentage rate of


12%, compounded monthly. How long will it take the investment to reach $2,000?

Answer: It would take about 5 years and 10 months for the investment to reach $2,000.

Explanation and Solution:

Step 1: The annual percentage rate is the rate that you would receive if the
interest were calculated at the end of the year. This means there was no
compounding during the year.
Step 2: Determine what the interest rate would be per month by dividing the 12%
by 12 months:

Step 3: From above, we know that we can find the balance after t years as
follows:
12

We use 12 t because there are 12 months in every year.

Step 4: Replace the right side of the above equation with $2,000:

Step 5: We must isolate the exponential term; therefore, divide both sides by
$1,000:

Step 6: Take the natural logarithm of both sides of the above equation:

Step 7: Simplify the left side of the above equation:

Step 8: We get:

Step 9: Note that 5.80505974113 years can be written 5 Years + 0.80505974113


years. If you multiply 0.80505974113 years by 12, you get 9.6607 months. This
indicates that it takes 5 years and about 10 months for the $1,000 to reach $2,000.
13

Applications to Mortgages

There is a relationship between the mortgage amount, the number of payments, the
amount of the payment, how often the payment is made, and the interest rate. The
following formulas illustrate the relationship:

where P = the payment, r = the annual rate, M = the mortgage amount, t = the number of
years, and n = the number of payments per year.

Example: Suppose you wanted to take out a mortgage for $75,000 with monthly
payments at 7%, but you could only afford $450 per month payments. How long would
you have to make payments to pay off the mortgage, and how much interest would you
pay for this payment period?

Answer: It would take you 616.10755485 months to pay off the mortgage, that is,
roughly 51 years and 5 months. The bank would either have you pay 616 payments of
$450 per month and one last payment of $48.40 or your 616th payment would be
$498.40.

Solution and Explanations:

Step 1: In the equation

substitute $75,000 for M (the mortgage amount), 7% for r (the annual interest
rate), 12 for n (the number of payments per year, and $450 for P (the mortgage
payment). You are solving for t (the number of months that you must make
payments)
14

Step 2: Multiply both sides of the above equation by 12:

Step 3: Multiply both sides of the above equation by

to get:

Step 4: Divide both sides of the above equation by 5,400:

Step 5: Subtract 1 from both sides of the above equation:

Step 6: Multiply both sides of the equation by - 1:

Step 7: Take the natural log of both sides of the above equation:

Step 8: Simplify the left side of the above equation using the third rule of
logarithms:

Step 9: Divide both sides by Ln(1.00583333333) and simplify:


15

Step 10: Divide both sides by - 1:

Step 11: The 12t stands for the number of months in t years or 616 months plus
part of a month. The number of years is 51.3422962552 or a little more than 51
years to pay off the mortgage.

The amount that you will pay over this term is (616)($450) = $277,200
+1(.107555063)(450) or $48.39977835 rounded to $48.40. Normally, the bank
will add this extra amount on the 616th payment. Sometimes they will have a
617th pay of $48.40. The total amount paid over the mortgage period is
$277,248.40.

Everything over the $75,000 mortgage is interest. The interest you will pay over
the 51 plus years is $277,248.40 - $75,000 = $202,248.40
16

Applications to Seismology

Before we start, let's talk about earthquakes and how we measure their intensity.

In 1935, Charles Richter defined the magnitude of an earthquake to be

where I is the intensity of the earthquake (measured by the amplitude of a seismograph


reading taken 100 km from the epicenter of the earthquake) and S is the intensity of a
''standard earthquake'' (whose amplitude is 1 micron =10-4 cm).

The magnitude of a standard earthquake is

Richter studied many earthquakes that occurred between 1900 and 1950. The largest had
magnitude of 8.9 on the Richter scale, and the smallest had magnitude 0. This
corresponds to a ratio of intensities of 800,000,000, so the Richter scale provides more
manageable numbers to work with.

Each number increase on the Richter scale indicates an intensity ten times stronger. For
example, an earthquake of magnitude 6 is ten times stronger than an earthquake of

magnitude 5. An earthquake of magnitude 7 is times strong than an

earthquake of magnitude 5. An earthquake of magnitude 8 is


times stronger than an earthquake of magnitude 5.

Example: Early in the 20th century the horrific earthquake in San Francisco registered 8.3
on the Richter scale. In the same year, another earthquake was recorded in South America
that was four times stronger. What was the magnitude of the earthquake in South
America? Let SF and SA denote San Francisco and South America, respectively.
17

Answer: The intensity of the earthquake in South America was 8.9 on the Richter scale.

Solution: Convert the first sentence to an equivalent mathematical sentence or equation.

Convert the second sentence to an equivalent mathematical sentence or equation.


18
19

Applications to Growth and Decay

Logarithms and Modeling

Many phenomena in nature seem to follow the law that an amount Y


varies with time according to the formula

Y=C ekt
where Y is the original amount (the amount at time t=0) and k is a
non-zero constant.

If k is positive, then the amount Y gets larger or grows with time, and
the amount is said to be experiencing exponential growth.
If k is negative, then the amount Y gets smaller or diminishes with
time, and the amount is said to be experiencing exponential decay.

When solving exponential growth/decay problems, two situations


often occur:

• Time is given; you are to find the amount at that given time.
This usually just involves evaluating the amount function.
• The amount is given; you are to determine at what time does
this amount occur. This usually involves solving an exponential
equation, which means logarithms will be needed.

Example: Carbon Dating

All living beings have a certain amount of radioactive carbon C14 in their bodies. When
the being dies the C14 slowly decays with a half-life of about 5600 years. Suppose a
human skeleton is found in Tahoe that has 42% of the original C14. When did the person
die?

Solution

We can use the exponential equation from above:

y = Cekt

After 5600 years there is C/2 C14 left. Substituting, we get:

C/2 = Cek(5600)
20

Dividing by C,

1/2 = e5600k

Take ln of both sides,

ln(0.5) = 5600k

so that

ln(0.5)
k = = -0.000124
5600

The equation becomes

y = Ce-0.000124t

To find out when the person died, substitute

y = 0.42C

and solve for t:

0.42C = Ce-0.000124t

Divide by C,

0.42 = e-0.000124t

Take ln of both sides,

ln(0.42) = -0.000124t

Divide by -0.000124

ln(0.42)
t= = 6995
-0.000124

The person died about 7,000 years ago.


21

Applications to Regression

Suppose P changes from 50 to 60. Then Log P changes from 3.9 to 4.1 = 0.20.

What is the relative change in P itself? It is (60-50) / 50 = 0.20.

That is remarkable! We have:

Change in Log P is approximately the relative change in P itself

Example: Consider the following demand equation.

Q = APb

Take logs and simplify:

Log Q = Log A + b Log P

For notational consistency, let Log Q = Y, Log A = b0 and Log P = X. We have

Y = b0 + b X

Suppose you ran a regression with data and found that b0 = 5.1 and b = 2.0. Writing
back in log form, we have:

Log Q = 5.1 + 2.0 Log P

Now, if Price increases by 1%, by how much does Q change?

The above rule gives us the answer immediately! The relative change in P is 1% = .01.
But this is also the change in Log P, from the above rule. Because the regression
coefficient is 2.0, the change in Log Q is 2 times 0.01 = 0.02. Since Log Q increases by
0.02, the rule assures us that Q itself has changed by .02. That is, quantity supplied
increases by 2% when price increases by 1%. The value 2% is also called “elasticity.”
22

Sound, Creation and Beethoven

What is the unit of measurement of sound? The answer is decibel. What is a decibel? It
is a logarithm of ratios!!!

Long, long ago, in an industry far, far away ... the decibel was born. The story goes
something like this: Telephone engineers discovered that a particular length of a
particular type of phone wire attenuated the wire’s signal power a certain amount. The
same length of the same type of wire always attenuated the input power by the same
ratio, regardless of the actual level of the input signal. They called this ratio the “bel”, in
honor of Alexander Graham Bell.

But the “bel” was much too cumbersome for dealing with small signal power ratios, so
the “bel” was divided by 10 and called the decibel. Mathematically, it works out

dB = 10log(P2/P1).

As you look at this formula, you’ll note that it does nothing more than describe a ratio
between two power levels, P2 and P1. In most cases, P1 is the input power to some
device, and P2 is its output power. That’s what decibels are all about: logarithm of ratios.

Example: Assume you have a 50-watt stereo and your neighbor has a 100-watt stereo.
How much more powerful, in decibels, is your neighbor’s stereo than yours? To find out,
plug the two stereo systems’ power levels into the previous decibel formula. I’ve re-
written it slightly to make it a bit easier to use.

dB = 10 x [log(P2/P1)]
dB = 10 x [log(100 watts/50 watts)]
dB = 10 x [log(2)]
dB = 10 x [0.301]
dB = 3.01

Your neighbor’s stereo is 3.01 dB more powerful than your stereo.

Example: If your favorite radio station decides to install a new transmitter that will
increase its output power from 10,000 watts to 20,000 watts, how much more powerful,
in decibels, is the new transmitter than the old one? Again, the answer can be found with
the decibel formula.

dB = 10 x [log(20,000 watts/10,000 watts)]


dB = 10 x [log(2)]
dB = 10 x [0.301]
dB = 3.01
23

In both cases the answer is 3.01 dB. How can that be? In the first example, the power
difference was 50 watts, and in the second example the power difference was 10,000
watts! How the heck can each one be equal to 3 dB? The answer is not related to the
absolute difference between two power levels, but rather the ratio of the two power
levels. In both examples the ratio is two. That is, one of the power levels is twice as much
as the other. It doesn’t really matter what the actual levels are, just the ratio between
them. That’s all decibels are: logarithm of ratios.

Creation

Most religious books have one common notion: the universe began with a sound. (Many
physicists also assume this: recall the theory of the Big Bang.) Thus, for example, the
Bible starts with: “In the beginning was the Word.” Likewise, in Hinduism, the Sanskrit
word “Ohm” is the sound of the universe at creation. But sound is measured as a
logarithm of ratios. And so you see why mathematics is truly the universal language!!
And why logarithms are so profound.

Beethoven

Ludwig Von Beethoven went stone-deaf half way into his adult life. That didn’t stop him
from composing the greatest symphony ever: The Ninth Symphony. Beethoven yearned
to hear any sound at all, but couldn’t. In his quest to understand sound without listening,
he wrote several masterpieces until he crafted the Ninth. He understood (more than any
other composer) that the human voice was by far the most beautiful instrument. And so
if you listen to the Ninth, in the very first measure or two the structure for the entire
symphony appears, particularly the last movement – the Ode to Joy. He transformed his
entire life’s pain to an everlasting gift to the human race in this last movement where he
celebrates the human voice. If you study the music sheet carefully you will see the
amazing genius work out scale transpositions unlike any other in music. Beethoven, like
many composers, composed on the pianoforte.

But what is a piano keyboard? Based on all of the above, you should now deduce that it is
really a logarithmic scale. Likewise, the human voice is a chain of frequencies, which are
measured in hertz, and which are functions of logarithms.