Transcript Slide 1
The Game of Algebra or The Other Side of Arithmetic Lesson 20 by Herbert I. Gross & Richard A. Medeiros © 2007 Herbert I. Gross next More on Inverse Functions next © 2007 Herbert I. Gross The concept of functions, including inverse functions and composition of functions, gives us an effective way to explain the informal process of “undoing” in a more mathematically precise way. For example, we have viewed subtraction as being “unaddition”. The more formal view is to refer to subtraction as being the inverse of addition. There is a strong connection between the word “inverse” as it is used here and the word “inverse” as it is used in the term “inverse function”. next © 2007 Herbert I. Gross For example, to “undo” adding 3 we subtract 3. That is: if we add 3 to a given number and then subtract 3 the result is the given number. In less formal terms, “we are back to where we started from”. In the language of functions, the instruction “Add 3” can be represented by the function f where f(x) = x + 3. The instruction “Subtract 3” can be represented by the function g where g(x) = x – 3. So if our input is 7, we see that f(7) = 10, and g(f(7)) = g(10) = 7. In a similar way, we see that g(7) = 4 and, f(g(7)) = f(4) = 7. next © 2007 Herbert I. Gross There was nothing special about our choice of 7 as our input. That is: with f and g defined by f(x) = x + 3, and g(x) by g(x) = x – 3; for any number x, f(g(x)) = g(f(x)) = x. More specifically… f(g(x)) = f (g(x)) (x – 3) = (x – 3) + 3 = x and… g(f(x)) = g (x (f(x)) + 3) = (x + 3) – 3 = x next © 2007 Herbert I. Gross The composition of these two functions, g and f, is itself a function, and it is a function that basically “does nothing”. That is: if we let I denote fog (or in this case, equivalently gof) we see that I(x) = x. For this reason, we refer to I as being the identity function. © 2007 Herbert I. Gross next Notes Just as multiplying a number by 1 doesn’t change the number, composing a function with I doesn’t change the function. That is, for any function, f … x ) I( f(x) ) = ( f(x)) and f( I(x) ) = f( I(x) Definition If f(g(x)) = g(f(x)) = I(x) = x for all x in the domain of f and the domain of g, we call f and g inverses of one another and next write f = g-1 and g = f-1. © 2007 Herbert I. Gross In essence, we obtain the inverse of a function (if there is an inverse function) by interchanging the input and the output (that is, the domain and the image). For example, suppose that… f(x) = x + 3 If we interchange x and f(x) the equation becomes… x = f(x) + 3 and if we subtract 3 from both sides of our equation we obtain… f(x) = x – 3 next © 2007 Herbert I. Gross The “tricky part” is that f(x) in the equation f(x) = x – 3 is not the f(x) in the equation f(x) = x + 3; rather it is its inverse because we interchanged the roles of x and f(x). In other words… If f(x) = x + 3, then f-1(x) = x – 3 The confusion results from the tradition that we usually denote the input of a function by x. To avoid this confusion we may let y = f(x) in the equation f(x) = x + 3 to obtain… y = f(x) = x + 3 …and if we then solve for x in the above equation we see that… next x = y – 3 = g(y) = f-1(y) © 2007 Herbert I. Gross Not every function has an inverse. For example, suppose we define a function on the set of all 3-digit numbers such that the output of the function is the sum of the 3 digits. Thus, for example, f(123) = 1 + 2 + 3 = 6 However, there are several other 3-digit numbers the sum of whose digits is 6. They are… 105, 114, 123, 132, 141, 150, 204, 213, 222, 231, 240, 303, 312, 321, 330, 402, 411, 420, 501, 510 and 600. So for example, 123 ≠ 510, yet f(123) = f(510) = 6. © 2007 Herbert I. Gross next However, if the function had had an inverse, it would mean that for each output there would have been only one input that yielded that output. In more mathematical terms, it means that if a ≠ b, f(a) ≠ f(b). Thus, in this case f-1 doesn’t exist. If m ≠ 0, the linear function f, defined by f(x) = mx + b always has an inverse. Namely, if m ≠ 0, its graph is either always rising or always falling. That is, if m > 0 the line is always rising, and if m < 0 the line is always falling. In either case, this means that no two inputs next can have the same output. © 2007 Herbert I. Gross To obtain the inverse of an “invertible” function, we interchange the input and the output. In terms of the graphs, we interchange the points (x,y) and (y,x). In general, these two points will be different. y For example, the point (3,5) is not the same as the point (5,3) (3,5) (5,3) x © 2007 Herbert I. Gross next In fact, the only time that these two points will actually be the same is when x = y. Hence, the points of intersection of the graphs of y = f(x) and y = f-1(x) will always be on the line whose equation is y = x. More specifically, the two graphs will be symmetric with respect to the line y = x. In other words, if we look at the graph y = f(x) through a mirror that is placed on the line y = x, the mirror image of this graph will be the graph y = f-1(x). © 2007 Herbert I. Gross next This is illustrated below with the graphs y = x + 3, y = x – 3, and y = x. y=x+3 y=x (6,9) y=x–3 (3,6) (9,6) (0,3) (6,3) (-3,0) (3,0) (0,-3) next For further practice, let’s draw the line… y = 2x + 3 y = f(x) = 2x + 3, …and reflect it about the line, y = x, (3,9) y=x y = f-1(x) …to obtain the line, y = f-1(x), (0,3) (9,3) Note that… f-1(x) = x – 3 2 next (3,0) (-3,-3) Note If m is negative, a similar analysis holds. Namely if m is negative it means that as x increases, f(x) decreases. Hence again, no two different values for x can yield the same value for f(x). Geometrically, it means that any horizontal line intersects the straight line that represents the graph of f at one and only one point. © 2007 Herbert I. Gross next Generalization If a function is continuous (that is, if its graph is “unbroken”) then it has an inverse if and only if its graph is either “always rising” or “always falling”. If its graph is “always rising” we say that the function is monotonically increasing; and if its graph is “always falling” we say that the function is monotonically decreasing. © 2007 Herbert I. Gross next Key Point In the language of functions we say that f is monotonically increasing, if whenever a > b, then f(a) > f(b) In other words: as x increases f(x) also increases ... And we say that f is monotonically decreasing if whenever a > b, then f(a) < f(b). That is, as x increases f(x) decreases. © 2007 Herbert I. Gross next For example, with respect to the function f where f(x) = 2x, notice that as x increases so also does f(x). In other words, the curve y = 2x which represents the graph of f is always rising. Hence, no horizontal line can intersect the graph at more than one point. © 2007 Herbert I. Gross next Key Point In geometric terms, our above discussion essentially means that if the curve that represents the continuous function is sometimes rising and sometimes falling, then the function will not have an inverse. Let’s illustrate this in the case for which f is defined by f(x) = x2. Its graph is the curve y = x2. Notice that while every vertical line intersects this curve at one and only one point, every horizontal line that is above the x-axis intersects the next curve in two places. © 2007 Herbert I. Gross Key Point In other words, if this function had an inverse, then for each output y there would be one and only one value of x for which (x,y) belongs to the graph. Since f(x) = x2 and since there are two values of x for which x2 = 4, it means that f-1 doesn’t exist. © 2007 Herbert I. Gross next Namely, the graph of f is the curve y = x2. y = x2 y (-3,9) (3,9) (-2,4) (2,4) (0,0) © 2007 Herbert I. Gross x next Notice that every vertical line intersects this curve at one and only one point. y This is illustrated for the line x = 2 which intersects the curve only at the point (2,4), and the line x = -2 which intersects the curve only at the point (-2,4) © 2007 Herbert I. Gross (-2,4) y = x2 (2,4) x x = -2 x=2 next On the other hand, if instead, we start with the horizontal line y = 4, we see that it intersects the curve at two different points, namely (2,4) and (-2,4). © 2007 Herbert I. Gross y y = x2 y=4 (-2,4) (2,4) x next Aside At first glance it may seem that not all vertical lines will intersect the curve y = x2. Notice, however, that no matter how large the number c is, the line y = c will intersect the curve y = x2 at the point (c,c2) © 2007 Herbert I. Gross next Although the line y = 4 intersects the curve in two points, notice that we can represent y = x2 as the union of the two curves y = x2 where x is non-negative, and y = x2 where x is negative. y (-3,9) (3,9) C2 C1 (-2,4) (2,4) (0,0) © 2007 Herbert I. Gross x next The curves C1 and C2 are the graphs of the functions g and h where g(x) = x2 when x is non-negative and h(x) = x2 when x is negative. In other words, if g and h are defined as above, g is a monotonically increasing function and h is a monotonically decreasing function. Therefore, the inverses of both g and h exist. © 2007 Herbert I. Gross next In summary… y = h(x) y = g(x) y The line y = 4 intersects the curve C1 only at the point (2,4), and intersects the curve C2 only at the point (-2,4). © 2007 Herbert I. Gross (-3,9) (3,9) C2 C1 y=4 (2,4) (-2,4) (0,0) x next Independently of whether the graph of a function is a curve that is always rising or always falling we may still interchange its x- and y-coordinates. Whenever we do this, the resulting curve will be the mirror image of original curve with respect to the line y = x, but it might not represent a function. To make this idea more concrete let’s look specifically at the curve x = y2 which represents the mirror image of the curve y = x2 with respect to the line y = x. In terms of functions, this curve represents the function x = [f(x)]2. © 2007 Herbert I. Gross next Note In our “plain English” version when we define the function f by the equation x = [f(x)]2 we are defining the function implicitly. That is: we are not explicitly telling what the program is, but what we are saying is that the program has the property that for each input the square of the output is always equal to the input. © 2007 Herbert I. Gross next next Note The concept of implicit and explicit is not restricted to the study of functions. For example, when we say that a number is 4 less than 7, we have implicitly defined the number 3. © 2007 Herbert I. Gross next The Curve x = y2 Suppose we want to draw the graph that represents the equation… x = j(y) = y2 Recalling that y2 cannot be negative, we see from the above equation that for any value of y, x must be at least as great as 0. Hence, the graph exists only to the right of the y-axis. © 2007 Herbert I. Gross next The Curve x = y2 To locate points on the curve x = y2, it’s probably less cumbersome to pick values for y and then to solve for the corresponding value(s) of x. © 2007 Herbert I. Gross For example… y 0 1 -1 2 -2 3 -3 4 -4 y2 0 1 1 4 4 9 9 16 16 x 0 1 1 4 4 9 9 16 16 (x,y) (0,0) (1,1) (1,-1) (4,2) (4,-2) (9,3) (9,-3) (16,4) (16,-4) next From our chart we see that the graph of x = y2 is… y x = y2 (9,3) (4,2) x (0,0) (4,-2) (9,-3) © 2007 Herbert I. Gross next The Curve x = y2 The fact that for each (positive) value of x there are two corresponding values of y means that x = y2 is not the graph of a function. More explicitly, if we take the square root of both sides of the equation x = y2, we obtain the equivalent equation… y = ±√ x © 2007 Herbert I. Gross next “y = ±√ x ” is an abbreviation for… “y = +√ x or y = -√ x ” In terms of a graph… y y = +√ x (9,3) (4,2) x (0,0) (4,-2) (9,-3) © 2007 Herbert I. Gross y = -√ x next While the line x = 4 intersects the curve x = y2 at both (4,2) and (4,-2); it intersects “y = +√ x only at (4,2) and y = -√ x ” only at (4,-2). y x=4 (9,3) (4,2) yx == y+√2 x x (0,0) (4,-2) (9,-3) © 2007 Herbert I. Gross y = -√ x next More generally, we see that the curve D1 defined by y = +√ x is always rising. Hence, it represents a function which we will denote by kp. That is: kp(x) = +√ x In a similar way, we see that the curve D2 defined by y = -√ x is always falling. Hence, it represents a function which we will denote by kn. That is: kn(x) = -√ x © 2007 Herbert I. Gross next In Summary… What we have shown thus far is that the function f, defined by f(x) = x2 does not possess an inverse. However, it is the union of two functions, g and h, both of which have an inverse. More specifically… g(x) = x2 when x is non-negative and h(x) = x2 when x is negative © 2007 Herbert I. Gross next In Summary… g(x) = x2 when x is non-negative h(x) = x2 when x is negative g-1(x) = kp(x) = +√ x when x is non-negative h-1(x) =kn(x) = -√ x when x is negative © 2007 Herbert I. Gross next In terms of a graph… y=x (-3,9) y = h(x) C2 (-2,4) (0,0) (4,-2) (9,-3) D2 y = h -1(x) © 2007 Herbert I. Gross next C1 y = g(x) y=x (3,9) (-3,9) y = h(x) C2 y = h-1(x) (2,4) D1 (-2,4) (9,3) (4,2) (0,0) (4,-2) (9,-3) D2 y = h -1(x) © 2007 Herbert I. Gross next next An Historical Note Given a continuous curve drawn at random; it is very unlikely that it is either always rising or always falling. Therefore, if a continuous curve is drawn at random, it will probably not represent a function that possesses an inverse. For this reason, until relatively recently, it was common to talk about multi-valued functions. In that context if we let f(x) = x2, we would have said that f-1 existed as a multi-valued function; and the two functions kp(x) and kn(x) would have been referred to as single valued branches of f-1. next © 2007 Herbert I. Gross A Historical Note Nowadays, however, we use the word “function” to mean “single valued function”; and we use the term “relation” rather than “multi-valued function”. That is, we would refer to f(x) = ± √x as being a relation. The distinction between a relation and a multivalued function is not too important because every relation (that is, a multi-valued function) can be represented by the union of two or more functions, each of which has an inverse. This was illustrated in our above discussion when we talked about trying to find the inverse of f in the particular case where f(x) = x2. next © 2007 Herbert I. Gross An Enrichment Discussion Suppose we had “invented” Prelude addition but had not yet “invented” subtraction. In this case, we would know that the function f, defined by f(x) = x + 3 must have an inverse. Namely because no two different values of x have the same image, we see that once we know the value of x + 3, we can uniquely determine the corresponding value of x. In terms of a graph, we could construct f-1 by first drawing the line y = x + 3 and then reflecting it about the line y = x. The resulting line would be the graph of f-1. next © 2007 Herbert I. Gross An Enrichment Discussion Notice that since we can draw its graph, the function f-1 exists even if we don’t bother to give the function a specific name. However, as we all know, it turned out that we “invented” the minus sign and defined f-1 by f-1(x) = x – 3. Thus, to solve an equation such as 3 + x = 12 using a calculator, we would translate this indirect computation into a direct computation by rewriting it in the equivalent form 12 – 3 = x. next next © 2007 Herbert I. Gross An Enrichment Discussion Of course this discussion might seem boring to us because we’ve know for a long time, even if not in these exact words, that subtraction is the inverse of addition. next © 2007 Herbert I. Gross An Enrichment Discussion The above discussion applies almost verbatim to any function f, provided that f is either monotonically increasing or monotonically decreasing. In particular it applies to the monotonically increasing function f where f(x) = 2x. However, we are most likely not nearly as comfortable discussing exponential growth and its inverse as we are when we are discussing addition and its inverse. So let’s construct f-1, when f(x) = 2x. next © 2007 Herbert I. Gross An Enrichment Discussion By way of review, starting with f(x) = 2x we can choose various values for x and compute the ordered pairs (x,2x). In this way we see that the points (-1,1/2), (0,1), (1,2), (2,4), (3,8)... belong to the curve y = 2x (which represents the graph of f). By interchanging x and y we see that the points (1/2, -1), (1,0), (2,1), (4,2), (8,3)…belong to the curve that represents the graph of f-1. Or looking only at the geometric graph, we start with the curve y = 2x and reflect it, point by point, about the line y = x. In this way the curve that is thus generated next represents the graph of f-1 © 2007 Herbert I. Gross Pictorially… Thus, f-1 exists whether or not we give it an explicit name. It turns out that the name we give to f-1 in this case is f-1(x) = log2x f(x) = 2x y=x (3,8) (2,4) (1,2) (0,1) (-1,1/ ) 2 (2,1) (1,0) (1,-1/2) (8,3) (4,2) f-1(x) = log2x next next © 2007 Herbert I. Gross An Enrichment Discussion There was nothing special about choosing the base to be 2 and then showing that if f(x) = 2x, then f-1(x) = log2x. Thus, if we replace 2 by b, we obtain the more general result… if f(x) = 2 b x, then f-1( x ) = log b2 x Because of place value notation, we often choose b to be 10. In this case we omit the subscript and simply write log x rather than log10x. next © 2007 Herbert I. Gross An Enrichment Discussion For example… 23 =8 means the same thing as log28 = 3 52 = 25 means the same thing as log525 = 2 4-2 = 1/16 means the same thing as log41/16 = -2 103 = 1,000 means the same thing as log 1,000 = 3 Or with a shift in emphasis… log636 = 2 means the same thing as 62 = 36 log381 = 4 means the same thing as 34 = 81 log21/8 = -3 means the same thing as 2-3 = 1/8 log 10,000 = 4 means the same thing as 104 =10,000 next © 2007 Herbert I. Gross An Enrichment Discussion If we try to use the calculator to find the value of x if, say, x = log315, we discover that there is no log3 key. It would have been nice if there were such a key because we could then enter 15 and press the log3 key to obtain the value of x. However, as we discussed in our previous lesson on exponential functions (Lesson 19), most calculators have a “log x” key. Reminder… The relationship between 10x and log x is that if f(x) = 10x, f-1(x) = log x. next © 2007 Herbert I. Gross Solving the Equation bx = c Where b and c Are Positive Constants 0.8451 7 log x Let’s use a calculator that has a “log x” key to solve the equation 10x = 7, we enter 7 on the calculator and then press the “log x” key. In this way, we obtain the result that to 4 decimal place accuracy, x = 0.8451. In other words, 100.8451 = 7. next © 2007 Herbert I. Gross The major problem is… What happens when the base is not 10? For example, suppose we want to find the value of x for which 7x = 28. Knowing that 7 =(10 100.8451,) we may replace the equation… 7x = 28 by the equivalent equation… 7x = 28 and by our rules for exponents, we can replace the above equation by the equivalent equation… 100.8451x = 28 © 2007 Herbert I. Gross next Solving the Equation bx = c 100.8451x = 28 Since log 28 means the power to which 10 has to be raised to obtain 28 and since the above equation tells us this power is 0.8451x 0.8451x, we see that the above equation may be rewritten as… 1.447158 log 28 = 2 Using our calculator we see that… 8 log x © 2007 Herbert I. Gross log 28 = 1.447158 next Solving the Equation bx = c So replacing log 28 by 1.447158 in the equation… log 28 = 0.8451x we obtain the equivalent equation… 1.447158 = 0.8451x And from the equation above, we see that… x = 1.447158 ÷ 0.8451 = 1.71259 next © 2007 Herbert I. Gross Checking the Solution for the Equation bx = c As a check we see that 71.71259 = 28.00957... Since 1.447158 = log log 77, log 28 28 and 0.8451 = log the derivation that took us from the equation 7x = 28 to the equation… x = 1.447158 ÷ 0.8451 can be rewritten as… If 7x = 28, x = 1.447158 ÷ 0.8451 next © 2007 Herbert I. Gross Solving the Equation bx = c If we replace 7 by b b… and 28 by y, y y… we obtain the more general result… If 7 x = 28, x = log 28 ÷ log 7 For further practice, let’s apply the above formula with b = 8 and y = 40 to find the value of x for which… 8x = 40 next © 2007 Herbert I. Gross Solving the Equation bx = c To Solve the Equation 8x = 40 In this case, the formula x = log y ÷ log b tells us that… x = log 40 ÷ log 8 Using the “log x” key we see that… log 40 = 1.60206 and log 8 = 0.903090 Therefore… X = 1.60206 ÷ 0.903090 = 1.7740 As a check we see that 81.7740 = 40.002 © 2007 Herbert I. Gross next This concludes our present discussion of inverse functions. Further practice is left for the Key Stone Illustrations and the Exercise Set for this Lesson. next © 2007 Herbert I. Gross