Calculus of One Real Variable

Prof. Joydeep Dutta

Department of Economic Sciences
Indian Institute of Technology, Kanpur

Lecture – 08
Intermediate Value Theorem

Welcome once again; as we continue moving towards more advanced things, do not get
so worried about the word advanced. It simply means that you have some more
additional information; that is what advancement means. Today we are going to speak
about a very important result, now many of you might think though mathematicians are
only bothered about some results and it has no consequence possibly in life, it might not
have always a consequence in practical life, but sometimes these results can be used to
prove some other results which might have consequences in your life; what does this
intermediate value theorem says? It essentially tries to tell you the following.

(Refer Slide Time: 01:13)

So, suppose if we have a function f[a,b]  R, right; assume that it is continuous, once
you assume that you have a you know huge; you have huge information of property of
the function, the nature and then it is says that suppose you have the following
information, this is a given information that f(a) is strictly bigger than 0, f(b) strictly less
than 0.
Then there exists; this is the symbol for this opposite; capital E is the symbol for there
exists x in a, b open interval a, b such that f(x) = 0; obviously, this x cannot be any of the
n point because n points are not 0. So, this is what is called the intermediate value
theorem, does there exists 1 x only or does there exists more than 1 x. Let us draw some
pictures to confirm it, for example, if you takeout term like this, if you have a continuous
function like this between a and b. So, this is your a and this is your b; f(a) is positive
and f(b) is negative and here is this point x; where it cuts the x axis means you start from
the top of mountain and go below the river, then you really have to cross the surface of
the river; once that is a basic idea, then that is basically it means you want to travel on
the number line from -2 to +2; you have to cross 0 once that that simple idea is in built
here.

Now, let us look at a situation where we can have more than one point where f; f(x)
would be 0. So, the function could be like this never mind you see f(a) is still bigger than
0 and f(b); this is b and this is a and f(b) is still less than 0, but you have 1, 2, 3, 4, 5
points at which the function attends the value 0. So, here the statement says that there
exists an x, it does not say that there is exists only 1 x. So, here I want to make this fact
clear that there can be more than 1 x possibly could be countably infinite or un countably
infinite xs. So, functions would look very complex we are not getting into that.

(Refer Slide Time: 04:42)

But this idea of continuity is very important; why I will tell you, for example, I will take
a function f(x) which I will define from 0 to 1 to R and I will define it as follows at f(x)
is equal to 1 when it is between 0 ≤ x ≤ 1/2 and is equal to -1 when it is between 1/2 ≤ x
≤1. So, you would you me immediately notice that here a is equal to 0 and b is equal to
1, a = 0 and b = 1 and what is f(a) at 0? F of 0 is 1 which is positive and f of 1 is minus 1
which is negative, but let us look at the function values. Let us draw the diagram, but you
can immediately see from the function; the values at the function is taking there is no x
for which the function is taken he x value 0 value why it has not; it is simply because of
the fact that in this line 0 to 1. So, at this is the point half. So, up to half the value; so,
this is the demarcating point. So, here the value was one, this is 1 and this is minus 1. So,
it comes here and here. So, at half exactly at half it takes the minus 1 value and not the 1
value.

So, this is the function value. So, you see immediately that there is a gap; this is a
discontinuous function and for a discontinuous function, the intermediate value theorem
need not hold true, it may hold true also I do not say it does not; for example, I will say
f(x) [-1,+1]  R, sorry to R and let me define it in this way; f(x) is equal to x when see
here; you must understand that I am always not restricting f(a) to be strictly bigger than 0
and f(b) to be strictly less than 0, it could be the opposite, it could be for example, that
f(b) is strictly bigger than 0 and f(a) is strictly less than 0 then also. So, here your either
f(b) to be strictly bigger than 0 and f(a) to be strictly less than 0, then also you would
have the same conclusion; there would be an x for which f of x would be equal to 0.

What now I am trying to show you that this conditions f a when that this f(a) and f(b) are
opposite of opposite sign. So, if I take the product of f(a) into f(b) that would be negative
that is just a sufficient condition and it is not a necessary condition for such a thing to
happen that whenever I find that there is a x between a and b such f(x) is equal to 0, it
does not mean that this is going to happen or the product of f(a) into f(b) would be
strictly bigger than 0. So, this is one form of the theorem to be more precise. Now you
can write it in a better form that if from a, b to R is continuous I am writing cont. for
short and f(a) into f(b) the product is negative, then there exists an x lie in the open
interval a, b such that f(x) must be equal to 0. So, this is the form which is much more
acceptable, right.
Now, here let me put f(x) is equal to x, when x is lying between 1 and 0 and f(x) equal to
-1, when it is lying between minus 1 and the interval minus 1, 0 excluding 0. So, this is
the functional value. So, if I draw this; what does this mean? So, here is my say -1. So,
here is my +1, here is 0. So, you come from here. So, in this case f(b) is +1, if I take b to
be plus 1 and a to be minus 1 and then from 0 to minus 1, I put the value minus 1. So,
here is -1.

So, this is my function, this is of course, a discontinuous function, it is not continuous,

but you can immediately see that the conclusion that f(x) = 0. So, here is an x between -1
and +1 for which f(x) = 0. So, this and also you can see the product f(a) which is 1, f(b)
is 1 and f of minus f of a which is minus 1 is minus 1. So, product is negative. So, this
what I am trying to say is that if this condition, for example, continuity has been failed,
but this is true then also this can remain true. So, these conditions that you see imposed
here that its continuous and this product is negative is a sufficient condition means if this
happens, then I am guaranteed that such a thing will definitely happens. So, this is
something you have to keep in mind.

These small examples actually give you much more better inside into what is really
going on and let us see, I will not immediately try to give a idea of proof. So, I am going
to do the proof of this because everybody might not be adopted to or getting used to
mathematical proofs though it is a necessary tool, but I would rather insist on concept
building here in this course which some cases coming to proofs, but I will try to give you
some first show some ways by which we can apply this result, let us see one of the first
conclusion consequence which I can draw from these intermediate value theorem.
(Refer Slide Time: 13:10)

So, this is what I would call in short the IVT. Now here is another form of intermediate
value theorem. So, this intermediate value theorem; if I go a little bit of the history was
first shown or rather proved by Bernard Bolzano and Bolzano by the way was a priest in
the beautiful city of the Prague which is now in the capital of Czeh Republic and believe
it or not; lot of very great works in science for example, has been done by priest or
monks because for very poor people at those times, monastery was one of the places
where they get an education they could get a food, etc. So, in Europe lot of very talented
people, but not from a rich background went to this monasteries and they really had they
had time to practice science and. So, this was this was, the contribution of Bernard
Bolzano. So, let us see; what is an immediate effect of that IVT.

So, may be first conclusion I should write corollary actually, but I am writing as a
theorem because this is what is written as IVT by many many books, but what I am
going to write down is IVT. So, let  f[a, b]  R be continuous, let c element of a, b
actually open interval a, b be such that c strictly lies between f(b) and f(a) you see it
could be opposite, it could be lie on f(b) and f(a) could be strictly bigger than f b here I
am just telling that let us take it like this then there would exists x  (a, b) such that f(x)
= c, never mind, I want to write tell you that this condition can also be written as it could
be of this form also, but just for the sake of showing how the results applies, we will just
consider this form; now how do we go around and really prove something of this nature.
So, in this case; the natural tricks. So, here is a trick which mathematicians are used what
I know is that there would be a point f(x) = 0, if I have 0; if I have the product of f a into
f b to be negative that is 0 should strictly lie between f(a) and f(b); 1 is negative, 1 is
positive then, but Neumann have see what I do is. So, if I want to prove it. So, consider a
new function g(x) g whose all values are evaluated like this f(x) - c, then let us look in to
what is g(a) = f(a) - c or what is f(a) – c, is negative because f(a) is strictly less than c.
Now what is g(b) = f(b) - c which is strictly positive because from this condition that f(b)
is strictly bigger than c. So, the, but because f is a continuous function; if you add a
constant to a continuous function, it remains a continuous function.

So, the first observation that we get from here is that g is a continuous function now what
is the product between g(a) and g(b) their product is strictly less than 0 because 1 is g(a)
is negative g(b) is positive now this means. So, hence now applying IVT there exists x in
this interval a, b such that g(x) is equal to 0 which would imply f(x) - c is equal to 0
which would imply f(x) is equal to x is equal to c. So, what are we proved that if you do
this; this is what going to happen you will go going if this is the condition by applying
IVT by doing this little trick a little translation, we can show just by a simple application
of that that there would always exists an x for which this happens.

Let us look into another interesting consequence of this, right. So, can I prove this fact,
every positive number has its square root, if I tell you prove to me that every positive
number has it is square root you would immediately try to take positive numbers and do
the square root and so common; it is obvious you can calculate the square root of
positive numbers what is so big about it, but if I tell you to demonstrate to me that every
positive number has a square root right, then which means that you have to
mathematically demonstrate that whatever positive numbers; I give to you there is a
number whose square is that number.

So, if I give you any number alpha you have to mathematically show the existence of an
x for which x of square is alpha it is not that just you will sit down and start writing those
you know start trailing on one has the square root 2 has the square root; it is not like that
in mathematics you do not do such exist or exhaustive things you can cannot do an
exhaustive checks because in calculus everything is infinite here this is the real of
infinite hence a mathematical demonstration is very important, but we will see how this
intermediate value theorem or even this one; if you want to say actually shows you that
you can really prove that every positive numbers has a square root. So, that is exactly
what I am going to show you.

(Refer Slide Time: 20:19)

So, there is a beautiful proof from the book of Spivak and so may be if I want to call the
theorem or mathematician would call corollary whatever you want to call it; it is a nice
result every positive numbers, every alpine that natural numbers set has a square root
every positive real number not in n, sorry, in R plus means in the positive part of R every
positive real number has a square root; say in order to prove this; what do I establish? I
have to establish the following fact that to establish. So, what is the thing that I have to
establish; I have to establish the following fact that given any alpha element of R and
alpha greater than 0, there exists another real number x whose square is this number. So,
x would be the square root of this number. So, there exists x element of R such that x
square of x is alpha which means x is the square root of alpha, right.

Now, let us write down the functions, here let me consider the function f(x) = x2 and
alpha is a given number to me. So, now, find b in R such that f(b) is strictly bigger than
alpha if alpha is strictly bigger than 1, you could take b equal to alpha because alpha
square root be strictly bigger than 1 and if you want if alpha is less than 1 and b to 0,
then you could take alpha equal to 1 because alpha square root be definitely less than 1
where alpha square would be less than alpha. So, whatever it is you can always find such
a number. So, here let me draw the diagram f(x) is equal to x square. So, here is my
alpha and here is my f(b) is my b and here is my f(b).

So, f b is strictly bigger than alpha, but alpha being a positive number is, but since alpha
is greater than equal to 0, it implies that alpha is strictly bigger than f 0 because f(0) = 0
here because it is x square f(0) is 0. So, what do I find f of 0 strictly less than alpha
strictly less than f(b), but in the interval 0 b, but in the f, if I just restrict it to the interval
0 b to R, then this is also a continuous function. Now once I do know that I know this has
happened; I can just apply; what I have learnt here with a = 0 and b = b there the same b.

So, this would imply that there exists an x in the open interval 0 b such that f(x) is equal
to alpha where this and this; obviously, implies where the f(x) is nothing, but x square, x 2
= α. So, you see how beautiful mathematics says that a. So, called advanced idea can is
actually needed to prove something which looks very obvious to you now we shall go
towards the understanding of the proof of this results how do you prove it actually, but
before doing.

So, I will tell you a story associated with this result there is a interesting story in many
books write this story, but here I will speak from a book called the concept of
mathematics by young steward about which I have already mentioned to you, but I still; I
would like to remind you and write down the name of this book and I expect that every
maths student would get an access to this every school library should have it maybe high
school level; not school, rather college library should have it in India.

Unfortunately not every; any college which can has money to buy books should have
these book at least.
(Refer Slide Time: 25:31)

It is a very cheap book at this at present, he is one of the most passionating writer in
mathematics; he is one of the leading mathematician over times, sorry, Ian Stewart from
Warwick university in the United Kingdom and this book is published by Dover
publications.

So, I begin by really showing you this book on the screen I hope you see that. So, please
consider if thus some even teachers watching this program please consider having this
book in your library, it is must for students and even infact, teachers to have a read of
have a reading of this book, math is not always about doing calculations or doing
manipulations math is about developing ideas brilliant mind developing ideas in biology
or mathematics is a brilliant mind, it is not that a mathematician who manipulates too
many things is much; he is much more intellectual faculty and than a biologist
developing a break through concept.

So, here is the puzzle, the puzzle is the following. So, I read out from this book which
says that at exactly 9 ‘o’ clock on Monday morning, a man sets of up the mountain
arriving at 6 ‘o’ clock in the evening. So, a man starts. So, here I can view on a
horizontal plane the topography of the mountain. So, here at 9 pm, he starts climbing and
6 pm, he reaches the top of the mountain and in the next day at 9 ‘o’clock next morning,
he descends from the same route arriving at a point which he set out at 6 ‘o’clock in the
evening.
So, 9 ‘o’clock next morning the graph would be like that at 9 ‘o’clock in the next
morning, he sets out like this and climbs back to this point. Now the puzzle is the
following prove that at some time he is at the same place both the days. So, there must be
some point here, say some point on the mountain route where he is present; he arrives at
the same time in both the days.

So, if he was at 12 ‘o’clock in some place on the mountain route while going up, he is at
12 ‘o’ clock at that same place while going down might looks; it might looks like dis-
strange to many of us; who has travelled up and down mountains the idea is the
following that. So, this guy is now climbing down the mountain in the morning. So,
assume that there is ghost of this man who is climbing up this mountain and he does the
opposite journey some man start 6 in the morning and does the opposite journey climbs
up the mountain. So, this is the path of this guy. So, he starts at 6 and this guy has started
at say 9 in the morning and he started climbing down.

So, another guy is actually now another guy has started at 6 and he is climbing up. So,
there is a time you see. So, here is up path here is a down path. So, in the down path what
the guy is doing he is climbing down you can think there is a ghost down path the; so, in
the down path this guy is coming down at starting at 9 and he is coming down, but think
that there is a ghost; who is also walking up the same path the next day towards the top
of the mountain, then you see that there is a time when both of their paths cross which
means the ghost and the man should be face to face at that point.

So, that this is nothing, but the intermediate value theorem and what is this form of
intermediate value theorem. So, we will just write down; how can I write this thing. So,
this is the particular way of talking about the intermediate value theorem. So, this is
something like this; see this curves are not continuous if the mountain profile was not
continuous such things would not happen as I already shown, but here the mountain
profile is also continuous. So, this is an application of the intermediate value theorem
that there exists a point between this where you get my whatever be the point there
would be a there would be a point where time point where this particular point would be
reached in both ways.

So, this can be actually expressed in the following way this idea this is the trick behind
proving the intermediate value theorem then I can write the intermediate value theorem
as the following. So, I left 2 functions f and g and they have very opposite behaviors. So,
behavior is this f(a) < g(a), but f(b) > g(b). So, this let me just draw the graph of this here
say if 2 functions say between a and b. So, here is my a and here is my b right, f(a) and
f(b). So, the function value at a; suppose this is my f, the graph of my function f is my f
function f(x) and incase of g you know that f, g(a) must be strictly bigger than f of a, but
g must be strictly lesser than f of b. So, this is my g.

So, there exists a point c between a and b where f(x) must be equal to g(x) and that is
exactly this puzzle right here because if you look at the mountain in this way it has one
functional graph if you look at the mountain in this way it is a another function of graph
and they have this same properties. So, this idea; so, what happens; now I consider a
function h whose h(x) is given by f(a) – g(a); right is given by f sorry, f(x) – g(x). So,
this function h is the continuous function between a to b because f and g are continuous
function between a to b and difference between 2 continuous functions is again a
continuous function.

So, this is a simple facts now what is h of a here h of a here is strictly less than 0 because
f of a is strict strictly less than g of a, but h of b is strictly bigger than 0 because f of b is
strictly greater than g of b which would imply again by using IVT that there would exists
an x in a, b such that h of x is equal to 0 which implies f(x) = g(x) at that point and that is
exactly the solution to this puzzle because while going up the mountain profile would
look as f while going down the mountain profile would look as g.

So, there is many-many ways to view this simple thing. So, how do you actually would
prove it I not really get into the I mean; idea the proof essentially is the following you
divide say the interval a, b into say 10 or 20 equal partisans equal small intervals look
into now for the first starting intervals you will observe because of the property of
continuity that if the difference of if h(a) < 0. So, in never hood h a would continue to
remain strictly less than 0. So, for sometimes you will have f(x) value strictly less than
g(x) values, but then look at the first interval where you find a value of x for which f(x)
value is bigger than g(x) value keep on dividing. Now divide that interval again and
among that interval first find. So, first find the interval where again f value is less than g
value.
So, you make these intervals smaller and smaller and smaller, right. So, so you in this
interval first find those values for which f means your you are not divided equally again
find the case where you first find an x for which f(x) > g(x). So, take that small, make
take another take that small interval equal on both sides and then again divide that. So,
you will keep on having this smaller intervals and smaller intervals basically; you will
taking intersectional intervals say; we will have smaller intervals and smaller intervals
and smaller intervals and which will silently close in on the point c at which f(x) is equal
to g(x).

So, this is the idea at which we have not getting in to, but one application the last
application a very important application is the idea which we will read; we now talk to
you now and that will be the last thing that I speak here and then the application of
intermediate value theorem is called the fixed point theorem. Fixed point theorem tells
me suppose I have a function of this nature that f: [0; 1]  [0,1] and is a continuous
function. So, here is my [0,1] on the x interval here is my 0 one interval on the y interval,
it will be slightly come this side. So, basically I have this square. Now I have some
continuous function some f is continuous insist domain is [0, 1], co domain is also [0,1];
what I am trying to say is that the fixed point theorem says is that there would exists an x
element of 0 one such that f(x) = x.

Now, of course, this function could have x 0 equal to 0. So, it is done or it could have f 1
equal to 1. So, it is done the question arises if not if f(0) is not and f(1) is not one the
function should always have a positive values. So, this whenever a point remains fixed
under the application of the function such a point x is called a fixed point. So, what does
it got to do with the intermediate value theorem, but look at what we have just learnt
consider my function f(x) is equal to f and consider another function g(x) is equal to x
f(x) is the one which I have given and considering the function g(x) is equal to x. Now f
0; if the f(0) is not equal to 0, then f(0) must be strictly bigger than 0 because the
function has to lie between 0 and 1, it cannot be negative.

So, f(0); so, the function value f(0) must be strictly bigger than 0, right that is f 0 is
strictly bigger than g(0). This implies this and what is f(1)? F of 1 cannot be more than 1;
f of 1 has to be strictly less than 1, either it is equal to 1, then it is done, if it is not equal
to 1, then it has to be strictly less than 1. The function value must be within 0 and 1 and
in that particular case, I just rub this part. So, just do it for you again. So, what is
happening here? So, I have taken f(x) as the function f is the function functional values
of f and for me g(x) = x.

Now, let me say that f of 0, let me be in a situation where f(0) is not equal to 0 and f 1 is
not equal to 1 because if any of these conditions happens then this result is true, then in
this case f(0) must be strictly bigger than 0 which implies that f(0) is greater than g(0),
then f(0) is not equal to 0 and the function is between the 0 and 1 which is non negative.
So, it has to be strictly bigger than 0, but f of 1 has to be now strictly less than 1 because
the function value cannot exceed 1 and f 1 is not equal to 1 which implies that f(1) < g(1)
which is 1 because g(x) = x.

So, this is exactly in this situation it is exactly in the situation of this mountain climbing
situation. So, you will immediately prove that there would be an x between 0 to 1 such
that such that f(x) = g(x), but g(x) = x. So, this would imply that f(x) is equal to x and
voila the theorem is proved, voila actually the French word which means that is it done.
So, if you see how a built up the whole story and surrounded with this simple man
climbing of the mountain which can come and prove this fact; it is; it looks very obvious
because you can draw this line y equal to x here and a continuous function you would
feel cut across this y equal to x line if it is; if the function range is also within 0 and 1.

Now, fixed point theorem you might say what is what story you are talk talking about,
but let me tell you a huge amount of modern economic theory and economics is of
course, you know practical stuff it is based on fixed point theorems. So, with this, I
would like to end this class and I hope you had enjoyed it.