Difference between revisions of "2021 AIME II Problems/Problem 13"
MRENTHUSIASM (talk | contribs) m |
MRENTHUSIASM (talk | contribs) m |
||
Line 6: | Line 6: | ||
<math>1000</math> divides this expression iff <math>8</math> and <math>125</math> both divide it. It should be fairly obvious that <math>n \geq 3</math>; so we may break up the initial condition into two sub-conditions. | <math>1000</math> divides this expression iff <math>8</math> and <math>125</math> both divide it. It should be fairly obvious that <math>n \geq 3</math>; so we may break up the initial condition into two sub-conditions. | ||
− | (1) <math>5^n \equiv n \pmod{8}</math>. Notice that the square of any odd integer is <math>1</math> modulo <math>8</math> (proof by plugging in <math>1^2,3^2,5^2,7^2</math> into modulo <math>8</math>), so the LHS of this expression goes <math>5,1,5,1,\ | + | (1) <math>5^n \equiv n \pmod{8}</math>. Notice that the square of any odd integer is <math>1</math> modulo <math>8</math> (proof by plugging in <math>1^2,3^2,5^2,7^2</math> into modulo <math>8</math>), so the LHS of this expression goes <math>5,1,5,1,\ldots</math>, while the RHS goes <math>1,2,3,4,5,6,7,8,1,\ldots</math>. The cycle length of the LHS is <math>2</math>, RHS is <math>8</math>, so the cycle length of the solution is <math>\operatorname{lcm}(2,8)=8</math>. Indeed, the <math>n</math> that solve this congruence are exactly those such that <math>n \equiv 5 \pmod{8}</math>. |
− | (2) <math>2^n \equiv n \pmod{125}</math>. This is extremely computationally intensive if you try to calculate all <math>2^1,2^2,\ | + | (2) <math>2^n \equiv n \pmod{125}</math>. This is extremely computationally intensive if you try to calculate all <math>2^1,2^2,\ldots,2^{100} \pmod{125}</math>, so instead, we take a divide-and-conquer approach. In order for this expression to be true, <math>2^n \equiv n \pmod{5}</math> is necessary; it shouldn't take too long for you to go through the <math>20</math> possible LHS-RHS combinations and convince yourself that <math>n \equiv 3 \pmod{20}</math> or <math>n \equiv 17 \pmod{20}</math>. |
With this in mind we consider <math>2^n \equiv n \pmod{25}</math>. By the Generalized Fermat's Little Theorem, <math>2^{20} \equiv 1 \pmod{25}</math>, but we already have <math>n</math> modulo <math>20</math>. Our calculation is greatly simplified. The LHS cycle length is <math>20</math>, RHS cycle length is <math>25</math>, the lcm is <math>100</math>, in this step we need to test all the numbers between <math>1</math> to <math>100</math> that <math>n \equiv 3 \pmod{20}</math> or <math>n \equiv 17 \pmod{20}</math>. In the case that <math>n \equiv 3 \pmod{20}</math>, the RHS goes <math>3,23,43,63,83</math>, and we need <math>2^n \equiv n \equiv 2^3 \pmod{25}</math>; clearly <math>n \equiv 83 \pmod{100}</math>. In the case that <math>n \equiv 17 \pmod{20}</math>, by a similar argument, <math>n \equiv 97 \pmod{100}</math>. | With this in mind we consider <math>2^n \equiv n \pmod{25}</math>. By the Generalized Fermat's Little Theorem, <math>2^{20} \equiv 1 \pmod{25}</math>, but we already have <math>n</math> modulo <math>20</math>. Our calculation is greatly simplified. The LHS cycle length is <math>20</math>, RHS cycle length is <math>25</math>, the lcm is <math>100</math>, in this step we need to test all the numbers between <math>1</math> to <math>100</math> that <math>n \equiv 3 \pmod{20}</math> or <math>n \equiv 17 \pmod{20}</math>. In the case that <math>n \equiv 3 \pmod{20}</math>, the RHS goes <math>3,23,43,63,83</math>, and we need <math>2^n \equiv n \equiv 2^3 \pmod{25}</math>; clearly <math>n \equiv 83 \pmod{100}</math>. In the case that <math>n \equiv 17 \pmod{20}</math>, by a similar argument, <math>n \equiv 97 \pmod{100}</math>. | ||
Line 57: | Line 57: | ||
We solve each congruence separately: | We solve each congruence separately: | ||
<ol style="margin-left: 1.5em;"> | <ol style="margin-left: 1.5em;"> | ||
− | <li>For <math>(1),</math> quick inspections produce that <math>5^1,5^2,5^3,5^4,\ | + | <li>For <math>(1),</math> quick inspections produce that <math>5^1,5^2,5^3,5^4,\ldots</math> are congruent to <math>5,1,5,1,\ldots</math> modulo <math>8,</math> respectively. More generally, <math>5^n \equiv 5 \pmod{8}</math> if <math>n</math> is odd, and <math>5^n \equiv 1 \pmod{8}</math> if <math>n</math> is even. As <math>5^n</math> is always odd (so is <math>n</math>), we must have <math>n\equiv5\pmod{8}.</math> <p> |
<i><b>That is, <math>\boldsymbol{n=8r+5}</math> for some nonnegative integer <math>\boldsymbol{r.}</math></b></i></li><p> | <i><b>That is, <math>\boldsymbol{n=8r+5}</math> for some nonnegative integer <math>\boldsymbol{r.}</math></b></i></li><p> | ||
<li>For <math>(2),</math> we substitute the result from <math>(1)</math> and simplify:</li> | <li>For <math>(2),</math> we substitute the result from <math>(1)</math> and simplify:</li> |
Revision as of 17:03, 8 September 2021
Contents
Problem
Find the least positive integer for which
is a multiple of
.
Solution 1
divides this expression iff
and
both divide it. It should be fairly obvious that
; so we may break up the initial condition into two sub-conditions.
(1) . Notice that the square of any odd integer is
modulo
(proof by plugging in
into modulo
), so the LHS of this expression goes
, while the RHS goes
. The cycle length of the LHS is
, RHS is
, so the cycle length of the solution is
. Indeed, the
that solve this congruence are exactly those such that
.
(2) . This is extremely computationally intensive if you try to calculate all
, so instead, we take a divide-and-conquer approach. In order for this expression to be true,
is necessary; it shouldn't take too long for you to go through the
possible LHS-RHS combinations and convince yourself that
or
.
With this in mind we consider . By the Generalized Fermat's Little Theorem,
, but we already have
modulo
. Our calculation is greatly simplified. The LHS cycle length is
, RHS cycle length is
, the lcm is
, in this step we need to test all the numbers between
to
that
or
. In the case that
, the RHS goes
, and we need
; clearly
. In the case that
, by a similar argument,
.
In the final step, we need to calculate and
modulo
:
Note that ; because
we get
.
Note that is
. We have
This time, LHS cycle is
, RHS cycle is
, so we need to figure out
modulo
. It should be
.
Put everything together. By the second subcondition, the only candidates less than are
. Apply the first subcondition,
is the desired answer.
~Ross Gao (Solution)
~MRENTHUSIASM (Minor Reformatting)
Solution 2
We have that , or
and
by CRT. It is easy to check
don't work, so we have that
. Then,
and
, so we just have
and
. Let us consider both of these congruences separately.
First, we look at . By Euler's Totient Theorem (ETT), we have
, so
. On the RHS of the congruence, the possible values of
are all nonnegative integers less than
and on the RHS the only possible values are
and
. However, for
to be
we must have
, a contradiction. So, the only possible values of
are when
.
Now we look at . Plugging in
, we get
. Note, for
to be satisfied, we must have
and
. Since
as
, we have
. Then,
. Now, we get
. Using the fact that
, we get
. The inverse of
modulo
is obviously
, so
, so
. Plugging in
, we get
.
Now, we are finally ready to plug into the congruence modulo
. Plugging in, we get
. By ETT, we get
, so
. Then,
. Plugging this in, we get
, implying the smallest value of
is simply
.
~rocketsri
Solution 3 (Chinese Remainder Theorem and Binomial Theorem)
We wish to find the least positive integer for which
Rearranging gives
Applying the Chinese Remainder Theorem, we get the following system of congruences:
It is clear that
from which we simplify to
We solve each congruence separately:
- For
quick inspections produce that
are congruent to
modulo
respectively. More generally,
if
is odd, and
if
is even. As
is always odd (so is
), we must have
That is,
for some nonnegative integer
- For
we substitute the result from
and simplify:
Note that
and
so we apply the Binomial Theorem to the left side:
Since
it follows that
![$\boldsymbol{r=5s+4}$](http://latex.artofproblemsolving.com/7/5/f/75f5b3b1c9f7b524ccdfd2471b0d473a60fe1567.png)
![$\boldsymbol{s.}$](http://latex.artofproblemsolving.com/b/c/9/bc9ff6f1810b9af8c4faac07244012d98b5cfb31.png)
Substituting this back into we get
As
is a multiple of
it has at least three factors of
Since
contributes one factor,
contributes at least two factors, or
must be a multiple of
Therefore, the least such nonnegative integer
is
Finally, combining the two results from above (bolded) generates the least such positive integer at
~MRENTHUSIASM (inspired by Math Jams's 2021 AIME II Discussion)
Solution 4 (Minimal Computation)
Note that
and
so
is periodic every
. Easy to check that only
satisfy. Let
. Note that by binomial theorem,
So we have
.
Combining
with
gives that
is in the form of
,
. Note that since
Easy to check that only
~Afo
See Also
2021 AIME II (Problems • Answer Key • Resources) | ||
Preceded by Problem 12 |
Followed by Problem 14 | |
1 • 2 • 3 • 4 • 5 • 6 • 7 • 8 • 9 • 10 • 11 • 12 • 13 • 14 • 15 | ||
All AIME Problems and Solutions |
The problems on this page are copyrighted by the Mathematical Association of America's American Mathematics Competitions.