Difference between revisions of "2020 INMO Problems/Problem 2"
(Created page with "==PROBLEM== Suppose <math>P(x)</math> is a polynomial with real coefficients, satisfying the condition <math>P(\cos \theta+\sin \theta)=P(\cos \theta-\sin \theta)</math>, for...") |
(→SOLUTION(1)) |
||
Line 18: | Line 18: | ||
Finally, we see that <math>P(x)=c+(1-x^2)^2h(x)</math> and <math>\text{deg} h<\text{deg} P</math> so <math>h</math> has the prescribed form. But then <math>P</math> also has the prescribed form, and our result follows. <math>\blacksquare</math> ~anantmdgal09 | Finally, we see that <math>P(x)=c+(1-x^2)^2h(x)</math> and <math>\text{deg} h<\text{deg} P</math> so <math>h</math> has the prescribed form. But then <math>P</math> also has the prescribed form, and our result follows. <math>\blacksquare</math> ~anantmdgal09 | ||
+ | |||
+ | <math>\textbf{Solution 2:}</math> | ||
+ | Let <math>\cos \theta + \sin \theta= c</math> and <math>\cos \theta - \sin \theta = d</math>. | ||
+ | Then <math>1-c^2=d^2-1</math>. | ||
+ | Let <cmath>P(x)=a_0+a_1(1-x^2)^2+a_2(1-x^2)^4+\dots+a_n(1-x^2)^{2n}</cmath> for all values of <math>x=c</math> and <math>x=d</math>. Then, clearly <cmath>P(c)=a_0+a_1(1-c^2)^2+ | ||
+ | ...a_n(1-c^2)^{2n}=a_0+a_1(d^2-1)^2+ | ||
+ | ...a_n(d^2-1)^{2n}=P(d)</cmath> as <math>1-c^2=d^2-1</math>. Thus, <cmath>P(x)=a_0+a_1(1-x^2)^2+ | ||
+ | ...a_n(1-x^2)^{2n}</cmath> for infinite values of x. Next, let <cmath>f(x)=P(x)-(a_0+a_1(1-x^2)^2+a_2(1-x^2)^4+\dots+a_n(1-x^2)^{2n}</cmath>) where <math>f(x)</math> must be a polynomial of degree <math>k</math>. | ||
+ | |||
+ | Then, <math>f(x)=0</math> for all values of x because a nonzero polynomial of degree <math>k</math> can have at most <math>k</math> roots, which is only possible when <cmath>P(x)=a_0+a_1(1-x^2)^2+a_2(1-x^2)^4+\dots+a_n(1-x^2)^{2n}</cmath> for all x. ~Faraz0007 |
Revision as of 14:47, 25 August 2025
PROBLEM
Suppose is a polynomial with real coefficients, satisfying the condition
, for every real
. Prove that
can be expressed in the form
for some real numbers
and non-negative integer
.
SOLUTION(1)
Assume to the contrary. Suppose satisfies
for all real
, and is of minimal degree and not of the prescribed form.
For some
, we have
.
Note that
. Set
. Then
as the latter vanishes at both
. Now let
for some
.
Then holds for all
, by plugging
in the original equation, since we have the identities
and
.
(Subtlety for beginners: while the equation in only holds for
away from roots of
, since these form a discrete subset of
, the equation extends to these as
is continuous.)
In particular, plugging we get
and
so
, hence
. Thus,
as desired.
Finally, we see that and
so
has the prescribed form. But then
also has the prescribed form, and our result follows.
~anantmdgal09
Let
and
.
Then
.
Let
for all values of
and
. Then, clearly
as
. Thus,
for infinite values of x. Next, let
) where
must be a polynomial of degree
.
Then, for all values of x because a nonzero polynomial of degree
can have at most
roots, which is only possible when
for all x. ~Faraz0007