Assignment 11
In the mid 1800’s Cauchy came up with the idea of using functional equations to describe the important functions of mathematics. The first case was linearity:
Definition 1 (Cauchy’s Functional Equation for Linearity) A function satisfies Cauchy’s functional equation if for all ,
Note it follows for any finite sum , one can ‘distribute’ a solution to Cauchy’s functional equation, by induction: .
Exercise 1 Prove that if is a continuous solution to Cauchy’s functional equation, then for some . That is, the continuous solutions to Cauchy’s Functional equation are precisely the linear functions.
Hint: follow the strategy below:
Let be a continuous function where .
- Prove that for all .
- Extend this to negative integers.
- Show that for *Hint: use that
- From the above, deduce that for rational , that .
- Now use continuity! If , then on …
Solution (Solution). Let . We will show that for all ,
Natural numbers
We prove by induction that for every ,
- Base case (): by definition.
- Inductive step: Assume . Then
2. Integers
First, so . For , hence Thus for all .
3. Reciprocals
Let . Since applying gives so
4. Rationals
Any rational with satisfies Hence for all .
5. All real numbers
Let . By density of , choose a sequence with . Continuity of gives
Therefore, the only continuous solutions of Cauchy’s functional equation are the linear functions
Exercise 2 Prove the exponential function for is monotone increasing: that is, if , .
Hint: we know its monotone on rational inputs, so the interesting cases are again when at least one is irrational (and, the argument for both irrational can be generalized to include the other case). Write down monotone increasing sequences, truncate the sequences until you can insure for all , and then apply the limit laws.
Solution (Solution). Let and consider the function . We will show that whenever in , one has
1. Monotonicity on rationals
First observe that if are rational numbers, then writing and with , , we have Since and exponentiation by an integer preserves order, it follows that implies . Hence is strictly increasing on .
2. Approximating by rationals
Now let be arbitrary real numbers. Choose two sequences of rationals such that and . One can form these by using density of rationals to pick and repsectively. Then by the squeeze theorem Because , we see that for all . (Alternatively, we could have chosen and both monotone increasing, and then found an where (by setting and applying the definition of convergence). Truncating the sequences at then gives the same result, that for all ).
3. Taking limits
By the rational‐case monotonicity, Both sequences and converge (by the continity of ) to and respectively. Hence by the limit laws for inequalities
Thus for any real , showing that is monotone increasing on .
Exercise 3 Prove range of the exponential function is all positive real numbers: that is, for any positive , show there is some where .
Hint: can you find some where ? If so, can you modify the idea to get an with ? Once you have these two values, can you apply a theorem about continuity?
Solution. Let and fix an arbitrary . We will show there exists such that
1. Finding an exponent above
Write with . By Bernoulli’s inequality, Since as , the Archimedean property guarantees we can choose so large that hence
2. Finding an exponent below
For any positive integer , Again by the Archimedean property we pick so large that thus $ a^{-M} < y.$
3. Applying the Intermediate Value Theorem
The function is continuous, and we have exhibited the existence of an with and an with . Thus on the closed interval there must be some where . Since was arbitrary, the range of is all of .
Exercise 4 Let be a logarithm and a rational number. Prove directly from the functional equation that .
Solution. Assume satisfies the logarithm functional equation
1. Value at
Set in the functional equation to get hence .
2. Positive integer powers
For any , note so by applying repeatedly,
3. Negative Integer Powers
Since and , we have so By replacing with , it follows that for every ,
4. Rational exponents
Let with and . By the definition of rational exponents, Applying gives But by the functional equation, so and hence
Thus for every rational the identity holds.
Exercise 5 Recall that an even function satisfies for all and an odd function satisfies .
Show that for differentiable functions, the derivative of an even function is an odd function, and the derivative of an odd function is an even function.
Solution. Let be an even function, so that for all . Taking the derivative of each side of this equality gives , and evaluating the right hand side with the chain rule: Thus, so is odd.
The same style of argument succeeds when starts out odd: now, differentiating the equality we see So, satisfies and the derivative is even.
Exercise 6 (Differentiating Sums) Let be functions which are both differentiable at a point . Prove that is also differentiable at , and
Solution. This calculation is directly analogous to derivative of done in the book: we use the limit laws to distribute the limit over addition, and note the result is the sum of two limit definitions of the derivative.
Exercise 7 Let be a function and be a point such that and is differentiable at . Prove that is also differentiable at and
Solution. Let , and let be differentiable at , lets start with . Then we can see Applying the limit here, we know that the second fraction is equal to , so we get,
Exercise 8 Prove the power rule, where is an integer
Hint: first do the positive case by induction; then use this to get the negative case.
Proof. We already know , so we can use this as our base case. To see how to proceed however we do the next case up as well. Using the product rule we see
Proceeding by induction, assume that for some this holds, and consider . Then
Hence the claim is true for all . For , note that by definition, and we can differentiate this identity using the power rule:
Simplifying: Solving this for yields
Thus the result holds for negative powers a well.