Homepage › Solution manuals › James Munkres › Topology › Exercise 7.9
Exercise 7.9
- (a)
- The formula
is not one to which the principle of recursive definition applies. Show that nevertheless there does exist a function satisfying this formula. [Hint: Reformulate so that the principle will apply and require to be positive.]
- (b)
- Show that the formula of part (a) does not determine uniquely. [Hint: If is a positive function satisfying , let for , and let .]
- (c)
- Show that there is no function satisfying the formula
Answers
(a) First, notice that does not satisfy the principle of recursive definition because, for , is not defined strictly in terms of values of for positive integers less than , since its definition depends on . Now we show that there does exist a function satisfying .
Proof. Consider the following reformulation:
where as a convention we take the positive square root for . Clearly for we have that is positive. Now suppose and that is positive for so that and are both positive. Then clearly is positive so that is defined and is positive. Hence is positive and well-defined for all by induction.
Thus, since depends only on values of for integers less than , this satisfies the recursion principle so that a unique satisfying the above exists. We also claim that this satisfies . Clearly, the explicitly defined values of and are satisfied. For , we have that so that, by definition,
which is the final constraint of so that it is also satisfied since was arbitrary. □
(b) First note that for the recursively defined function from part (a),
Now define the function as in the hint, that is for and . Then we clearly have while
so that for , and hence is violated. So it would seem that the hint as given does not exactly work.
Now we show that the function satisfying is not unique, taking inspiration from the hint.
Proof. We construct a function , different from from part (a), that also satisfies . We define using recursion:
Clearly, since each is defined only in terms of for (or without dependence on any values of ), exists uniquely by the recursion principle so long as each is well-defined. We show this presently by induction.
Clearly is defined for . For we have . Now, since , we have that so that and hence the square root, and therefore , is defined and positive. Now consider any and suppose that is positive. Then clearly is defined and positive since , noting that even if , its square is non-negative.. This completes the induction that shows that is uniquely defined.
Clearly since . Also obviously satisfies explicitly for . For we have
Then, for we have so that, by definition,
Thus the recursive part of holds for so that holds over the whole domain of as desired. □
(c)
Proof. Suppose that such a function does exist. Since the recursive property holds for , we have
Similarly, we have
so that either or . In either case, we have , which is of course impossible since the square of a real number is always non-negative! So it must be that such a function does not exist. □