Difference between revisions of "1973 IMO Problems/Problem 5"
(Created page with "<math>G</math> is a set of non-constant functions of the real variable <math>x</math> of the form <cmath>f(x) = ax + b, a \text{ and } b \text{ are real numbers,}</cmath> and <ma...") |
(→Solution) |
||
Line 12: | Line 12: | ||
==Solution== | ==Solution== | ||
+ | |||
+ | First, observe that for each function <math>f(x)=ax+b</math> in <math>G</math>, if <math>a=1</math> then <math>b=0</math>. This is a result of (c); for example, <math>f(x)=x+1</math> could not be in <math>G</math> because it does not have a fixed point. Or if <math>f(x)=x</math>, then every point is a fixed point. | ||
+ | |||
+ | Also, for each function <math>f(x)=ax+b</math> in <math>G</math>, if <math>a \neq 1</math> then the fixed point of <math>f</math> is where <math>y=ax+b</math> intersects <math>y=x</math>, namely where <math>x=\frac{b}{a-1}</math>. | ||
+ | |||
+ | Now, take <math>f_1(x)=a_1 x+b_1</math> and <math>f_2(x)=a_2 x+b_2</math>, both in <math>G</math>. By (a), <math>(f_1 \circ f_2)(x)=f_1(f_2(x))=f_1(a_2 x+b_2)=a_1 a_2 x + a_1 b_2 + b_1</math> and <math>(f_2 \circ f_1)(x)=f_2(f_1(x))=f_2(a_1 x+b_1)=a_1 a_2 x + a_2 b_1 + b_2</math> must also both be in <math>G</math>. By (b), <math>(f_1 \circ f_2)^{-1}(x)=\frac{x - a_1 b_2 - b_1}{a_1 a_2}</math> must also be in <math>G</math>. Finally, by (a), | ||
+ | |||
+ | <math>((f_1 \circ f_2)^{-1} \circ (f_2 \circ f_1))(x)=\frac{a_1 a_2 x + a_2 b_1 + b_2 - a_1 b_2 - b_1}{a_1 a_2}</math> | ||
+ | <math>= x + \frac{a_2 b_1 + b_2 - a_1 b_2 - b_1}{a_1 a_2}</math> | ||
+ | |||
+ | must also be in <math>G</math>. | ||
+ | |||
+ | Using our first observation, <math>\frac{a_2 b_1 + b_2 - a_1 b_2 - b_1}{a_1 a_2}=0</math>. Rearranging, we get <math>\frac{b_1}{a_1-1}=\frac{b_2}{a_2-1}</math>. Therefore, the fixed point of <math>f_1</math> equals the fixed point of <math>f_2</math>. Since we made no assumptions about <math>f_1</math> and <math>f_2</math>, this is true for all <math>f</math> in <math>G</math>. |
Revision as of 18:15, 25 October 2014
is a set of non-constant functions of the real variable of the form and has the following properties:
(a) If and are in , then is in ; here .
(b) If is in , then its inverse is in ; here the inverse of is .
(c) For every in , there exists a real number such that .
Prove that there exists a real number such that for all in .
Solution
First, observe that for each function in , if then . This is a result of (c); for example, could not be in because it does not have a fixed point. Or if , then every point is a fixed point.
Also, for each function in , if then the fixed point of is where intersects , namely where .
Now, take and , both in . By (a), and must also both be in . By (b), must also be in . Finally, by (a),
must also be in .
Using our first observation, . Rearranging, we get . Therefore, the fixed point of equals the fixed point of . Since we made no assumptions about and , this is true for all in .