Section 4.1: Introduction to Logarithms

In this section, we want to introduce the concept of and notation for a logarithmic function.  We will then discover the basic characteristics of a logarithmic graph and discuss how to create such graphs by hand.

Introduction to Logarithmic Functions

In the previous chapter, we examined exponential functions, such as f(x) = 2x.  We we graphed them as continuous functions, so that their range included all positive values.  That means that we should be able to solve equations, such as, f(x) = 32 or f(x) = 18, for x.  Using XFunctions, by selecting the function, f, you can see a plot of this function:

  

By clicking with the mouse, or experimenting with the "x=" box (or by guessing), you can easily find that f(5) = 32, so the only solution to 32 = 2x is x = 5.  However, it is much more difficult to solve 18 = 2x.  

With some effort, you might try to narrow down on a solution by taking a series of estimates:

x y = f(x) = 2x
4
4.1
4.2
4.15
4.18
4.17
·
·
·
24 = 16
24.1 = 17.1484...
24.2 = 18.3792...
24.15 = 17.7531...
24.18 = 18.1261...
24.17 = 18.0009...
·
·
·

While  x » 4.17 is a pretty good approximation to the solution, as you might expect, we can never write down all the correct digits of this number.  That is because, just like , p, and e, this is an irrational number.  

In the same way mathematicians were led to invent the square root function and the radical sign to allow us to write down the positive solution to 18 = x3 as , we will write the solution to 18 = 2x as log2(18).  We read "log2" as "log base 2".  Just as the cube root function is the inverse of the cube function, log2 is the name of the function which is the inverse of f(x) = 2x (i.e., the exponential function with base 2).  Note: f does pass the horizontal line test, and so is invertible.  While we began working with functions with nice algebraic formulas (e.g., x2 + 3x), we proceeded to give them one letter names (e.g., f(x) = x2 + 3x), and then names with more than one letter (e.g., id, abs, and int).  This notation is even more daunting, since we are using more than one letter (i.e., log) and a subscript (in this case, 2) to indicate the base.  Just as there are many exponential functions (e.g., 2x, 3x, (0.25)x, that is, one for each possible base), there are many different logarithmic functions, or "logarithms" for short, each of whose name is distinguished by a different subscript (e.g., log2(x), log3(x), log0.25(x)).  

Notice the asymmetry in the naming of the two groups of functions.  While we use a formula to describe exponential functions, we simply use a manufactured name to refer to logarithms.  The connection between the two groups of functions becomes clearer, if we introduce corresponding names for the exponential functions, such as exp2, exp3, and exp0.25, so that exp2(x) = 2x, exp3(x) = 3x, exp0.25(x) = (0.25)x.

Since logarithms are defined as inverse functions for exponentials, we know that they will cancel out with exponential functions.  Specifically:

For all numbers b > 0, and not equal to 1, logb(x) is the inverse of expb(x) = bx, so that they "undo" one another.  As an arrow diagram, we can indicate this as:

    or    

Algebraically, this means that logb(expb(x)) = x and expb(logb(y)) = y.  This looks more natural, if we use the traditional exponential notation, as logb(bx) = x and .  In particular, logb(y) is the unique solution to y = bx

Note: We have to exclude the base b = 1, since exp1(x) = 1x = 1 does not pass the horizontal line test.  

As we have seen, this is exactly the type of algebraic property that we need in order to solve equations involving exponentials.  In addition, we know that their graphs are simply reflections of one another.  We will discuss this more in the next section.

Just as, in general, exponential functions can only be computed using a very difficult formula from Calculus, logarithms require an equally difficult formulaUnfortunately, your calculator may very well not include a specific button for most logarithms!  This means that the only way for us to understand logarithmic functions is through their properties, namely, in terms of their inverse relationship to exponentials.  We will eventually use this relationship, and properties of exponentials to derive many useful algebraic properties of logarithms.  This will allow us to solve exponential equations and compute values for all logarithmic functions.

Practice using logarithmic functions to express the solutions to exponential equations by completing the following Exercises.  


Graphing Logarithmic Functions

Since we know that logarithms are inverse functions to exponentials, we can immediate construct their graphs by hand, and determine a number of their general properties.  For example, since log2 = exp2 -1 (where exp2(x) = 2x), we can simply take a table for values for exp2:

x y = exp2(x)
-¥
-1
0
1
0
1/2
1
2

to obtain a table of values for log2:

x y = log2(x)
0
1/2
1
2
-¥
-1
0
1

Since we know that the graph of exp2:

we can infer that the graph of log2:

Geometrically, by interchanging the roles of the horizontal and vertical axes, the graph of exp2:

becomes the graph of log2:

You can see that this graph does go though the three points, (1/2, -1), (1, 0), and (2, 1), from the table, and has a vertical asymptote corresponding the the "value" of -¥ at x = 0.  From this we can be describe the graph more naturally by saying that:

Notice how the shape of the general graph (increasing, curving downwards) is already suggested by our three points.  This mean that, when "connecting-the-dots", we only need to be careful to have the graph approach the vertical asymptote and go off to "infinity" on the right.

Since exponential graphs tend to grow very quickly (remember how they almost look like a right angle, when you "zoom out" far enough), logarithmic functions grow very slowly.  For example, even when x = 1,000,000log2(x) » 20.  This was the reason why they were first invented by John Napier in 1614.

Note: Using logarithms, Napier was able to convert the very large numbers involved in astronomical calculations into smaller, more manageable numbers (which helped Johannes Kepler to demonstrate his laws of planetary motion in 1628).  In fact, using logarithms, Napier was able to create one of the first primitive "calculators" out of a collection of small rods (which later were later called "Napier's bones").  William Oughtred improved on this to create the "slide rule" in 1632, which was used by scientists and engineers for over 300 years, until it was finally replaced in the 1970's by the first hand-held, digital calculators.

It is important to notice that, since we have already observed how all exponential graphs are related by a horizontal scaling and/or flip, it must be true that:

All logarithmic graphs are related by a vertical scaling and/or flip.  This means that any logarithm may be written as a multiple of any other logarithm.

We will explain this phenomenon in more detail in the next section, when we discuss the algebraic properties of logarithms.

In particular, this means that, since we now know how the graph of log2 looks, we have a pretty good idea how the graph of all logarithmic functions will look: 

You can quickly view a sampling of all types of logarithmic graphs, as the base ranges from b = .2 to 1.8 in the Example of the following applet: 

Note: You may have noticed that we had to express logk(x) in XFunctions as ln(x)/ln(k); we will explain this formula in the next section.  Notice how there is no graph when the base is 1, since log1 is not defined (i.e., 1x is not invertible).

You can create different examples of logarithmic graphs, by entering your own base in the following box and clicking the "Set Base" button:

Select the base, b =

This defines the functions "expb" and "logb" in XFunctions, corresponding to the exponential and logarithm, expb and logb, with the base you specify.  It also creates an Example, showing the two graphs on the same axes.  Try several examples and notice how the two graphs are always reflections of one another.

To summarize what we have seen so far, we can graph any logarithmic function, logb, as follows:

Since we can obtain a table of points and a rough sketch of the graph, we can then use our transformational graphing technique to construct plots of functions with a logarithmic "core" such as, f(x) = -2log0.4(x +1) + 3.  This decomposes as:

We begin with a table of values for exp0.4(x) = (0.4)x:

x y = exp0.4(x)
-1
0
1
¥
(0.4)-1 = 1/0.4 = 2.5
1
0.4
0

which gives a table of values for log0.4:

x y = log0.4(x)
2.5
1
0.4
0
-1
0
1
¥

Since b < 1, we know that this graph starts at y = ¥ at x = 0, decreases through the points, (0.4, 1), (1, 0), and (2.5, -1), while curving upwards:

From this, we can perform the following analysis:

Corresponding
Algebraic formula
log0.4(x) log0.4(x + 1)   2log0.4(x + 1) -2log0.4(x + 1) -2log0.4(x + 1) + 3
Geometric Effect Take the logarithm
function with base 0.4
Shift left 1 Stretch vertically
by a factor of 2
Flip vertically Shift up 3
Numerical Effect   Subtract 1 Apply the logarithm
function with base 0.4
Multiply by 2 Negate Add 3
Numerical 
Results
Inputs Outputs
2.5 1.5 -1 -2 2 5
1 0 0 0 0 3
0.4 -0.6 1 2 -2 1
0 -1 ¥

¥

We can imagine roughly how the graph will look:

Drawing in the asymptote, plotting the points and "connecting-the-dots" gives:

Practice graphing logarithmic functions by completing the following Exercises.  


Go to Properties of Logarithms


Table of Contents Send questions or comments to jwicks@northpark.edu Glossary