Set Theory is difficult, isn't it?

Logic Level 4

The set S S is formed according to the following rules:

1 ) 1) 2 S 2 \in S .

2 ) 2) If n n is in S S , then ( n + 5 ) (n+5) is also in S S .

3 ) 3) If n n is in S S , then 3 n 3n is also in S S .

Find the largest integer in the set 1 , 2 , 3 , , 2008 { 1,2,3, \ldots, 2008 } that does not belong to S S .


The answer is 2005.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

7 solutions

Kevin Klida
Apr 6, 2015

From 1) and 2) we get that every number greater than 2 that is 2 mod 5 is also in S.

Using 3) once tells us that 6 is in S since 2 * 3 = 6. This combined with 2) tells us that every number greater than 6 that is 1 mod 5 is also in S.

Using 3) again tells us that 18 is in S since 6 * 3 = 18. This combined with 2) tells us that every number greater than 18 that is 3 mod 5 is also in S.

Using 3) once more tells us that 54 is in S since 18 * 3 = 54. This combined with 2) tells us that every number greater than 54 that is 4 mod 5 is also in S.

One more application of 3) tells us that 162 is in S since 54 * 3 = 162. We already knew this, however, because 162 is greater than 2 and is 2 mod 5. Repeated applications of 3) from here yield no new elements of S.

This process tells us that every number that is greater than 54 that is either 1, 2, 3, or 4 mod 5 is in S. Therefore, our answer must be 0 mod 5.

It can be shown that no number that is 0 mod 5 is in S. Via Siddhartha Srivastava :

Let the smallest member of S divisible by 5 be x x . We either get x x from x 5 x - 5 or x 3 \frac{x}{3} . But both of these are smaller than x x and are divisible by 5, which is a contradiction, since x x is the smallest member divisible by 5. Therefore no member of S is divisible by 5.

This means that 2005 is our answer since it is the largest number in the given set of possible answers that is 0 mod 5.

Nice solution, but you make one recurring mistake.

Using 3) again tells us that every number that is 1 mod 5 is in S

Not true. For example, 1 1 isn't in S. Each number greater than 6 6 and that is 1 mod 5 is in S.

You make the same mistake for 3 mod 5 and 4 mod 5.

Also, typo in the 4 mod 5 case. It should be

This is because 3 * 3 = 9 which is 4 mod 5.

Siddhartha Srivastava - 6 years, 2 months ago

Log in to reply

Oh! Thanks for pointing out my mistakes. How careless of me. I've edited my solution appropriately (I hope!).

Kevin Klida - 6 years, 2 months ago

I think you gotta add up a little more - you should even prove that no number that is congruent to 0 in mod 5 is in S. Hint :- Look up sir Chew-Seong Cheong solution !

Venkata Karthik Bandaru - 6 years, 2 months ago

Log in to reply

That is a 2 line addition.

Let the smallest member of S divisible by 5 be x x . We either get x x from x 5 x - 5 or x 3 \frac{x}{3} . But both of these are smaller than x x and are divisible by 5, which is a contradiction, since x x is the smallest member divisible by 5. Therefore no member of S is divisible by 5.

Siddhartha Srivastava - 6 years, 2 months ago

Every element of S S can be expressed as a ( i , j ) = 2 ˙ 3 i + 5 j a (i,j) = 2\dot{} 3^i + 5j , where i , j = 0 , 1 , 2 , . . . . i,j = 0,1,2,....

Even for a ( m , n ) = 3 a ( i , j ) = 3 ( 2 ˙ 3 i + 5 j ) a(m,n) = 3a(i,j) = 3(2\dot{} 3^i + 5j) = 2 ˙ 3 i + 1 + 5 ( 3 j ) = a ( i + 1 , 3 j ) \quad \quad \quad \quad \quad \quad \space = 2\dot{} 3^{i+1} + 5(3j) = a(i+1,3j) .

We note that a ( i , j ) 2 ˙ 3 i + 5 j ( 5 + 1 ) 3 i 1 + 5 j 3 i 1 ( m o d 5 ) a (i,j) \equiv 2\dot{} 3^i + 5j \equiv (5+1)3^{i-1} + 5j \equiv 3^{i-1} \pmod{5} for i > 0 i>0 . Since 3 i 1 0 3^{i-1} \ne 0 , this means that multiples of 5 5 are not members of S S 5 = { 5 , 10 , 15 , . . . , 2005 } S\quad \Rightarrow \overline{S}_{5|} = \{ 5,10,15,...,2005 \}

We know that integer with units digit being 2 2 are S \in S , because, for k = 0 , 1 , 2 , . . . k=0,1,2,... , we have:

n 2 = 10 k + 2 = 2 ˙ 3 0 + 5 ( 2 k ) S for all k S 2 = { } n_2 = 10k + 2 = 2\dot{}3^0 + 5(2k) \in S \text{ for all }k\quad \Rightarrow \overline{S}_{2} = \{ \space \}

Similarly,

\(\begin{array} {} n_7 = 10k + 7 = 2+ 5 + 5(2k-1) & \Rightarrow \overline{S}_{7} = \{ \space \}\\ n_6 = 10k + 6 = 6 + 5(2k) & \Rightarrow \overline{S}_{6} = \{ \space \}\\ n_1 = 10k + 1 = 6+ 5 + 5(2k-1) & \Rightarrow \overline{S}_{1} = \{ 1 \}\\ n_8 = 10k + 8 = 18+ 5(2k) & \Rightarrow \overline{S}_{8} = \{ 8 \}\\ n_3 = 10k + 3 = 18+ 5 + 5(2k-1) & \Rightarrow \overline{S}_{3} = \{ 3,13 \} \\ n_4 = 10k + 4 = 54+ 5(2k) & \Rightarrow \overline{S}_{4} = \{ 4,14,24,34,44 \}\\ n_9 = 10k + 9 = 54+ 5 + 5(2k-1) & \Rightarrow \overline{S}_{9} = \{ 9,19,29,39,49 \} \end{array} \)

S = S 0 S 1 S 2 S 3 S 4 S 5 S 6 S 7 S 8 S 9 = { 1 , 3 , 4 , 5 , 8 , 9 , 10 , 13 , 14 , 15 , 19 , 20 , 24 , 25 , 29 , 30 , 34 , 35 , 39 , 40 , 44 , 45 , 49 , 50 , 55 , 60 , . . . , 1995 , 2000 , 2005 } \overline{S} = \overline{S}_0 \cup \overline{S}_1 \cup \overline{S}_2 \cup \overline{S}_3 \cup \overline{S}_4 \cup \overline{S}_5 \cup \overline{S}_6 \cup \overline{S}_7 \cup \overline{S}_8 \cup \overline{S}_9 \\ \quad = \{ 1,3,4,5,8,9,10,13,14,15,19,20,24,25,29,30,34,35,39, \\ \quad \quad \quad 40,44,45,49,50,55,60,...,1995, 2000, 2005 \}

Therefore the largest integer that does not S \in S is 2005 \boxed{2005} .

Sir, there is a lot more elegant solution to this problem than this ! 😓 😱 😥

Venkata Karthik Bandaru - 6 years, 2 months ago

Log in to reply

Okay, I was thinking but could not come out with one. I suppose it has something to do with multiple of 5,

Chew-Seong Cheong - 6 years, 2 months ago

Log in to reply

Sir please reshare this problem so that we can get more responses ! This has an elementary and really elegant solution 💎 !

Venkata Karthik Bandaru - 6 years, 2 months ago

Log in to reply

@Venkata Karthik Bandaru Can we have any more solutions??

Writabrata Bhattacharya - 6 years, 2 months ago

Perfect Solution now sir !

Venkata Karthik Bandaru - 6 years, 2 months ago

Unit digits of first few numbers. For (n+5):- 2 , 7 , 2.... r e p e a t s \color{#D61F06}{ 2, 7, 2....repeats} .
For 3n:- 2, 6, 18, 54, 162 >>>- 2 , 6 , 8 , 4 , 2.... r e p e a t s \color{#D61F06}{ 2, 6, 8, 4, 2....repeats}
Applying (n+5) on these, :-7, 11, 13, 9>>>> 7 , 1 , 3 , 9 \color{#D61F06}{7, 1, 3, 9} .
So all 1, 2, 3, 4, 6, 8, 9 unit digits greater than 54 are covered by (n+5) or 3n.
In fact 3n only supply seed of 6, 8, and 4 for numbers greater than 54.
It is (n+5) that covers all digits except 0 and 5 for numbers greater than 54.
So the highest number is 2005. Second highest is 2000.




Siddharth Iyer
Apr 6, 2015

We note that each element is of the form 2 (3^a)+5y. Proof; We definitely know that each element is of the form (2+5k) (3^a) +5z= 2*(3^a) + multiple of 5 as required. There is one thing an element will never be... a multiple of 5. Hence 2005 is the largest answer.

You have established that every element in S will never be a multiple of 5. It doesn't completely tell us that 2005 is the largest number not in S ! There are other possibilities greater than 2005 you can pick in that case. Just because you claim that no multiple of 5 is not in S, it doesn't mean that 2006, 2007 and 2008 are in S. You must explain why 2006,2007 and 2008 are in S.

Venkata Karthik Bandaru - 6 years, 2 months ago

Log in to reply

Yes you are right I should have fixed 2005 and checked numbers that are greater than it; 2006, 2007 and 2008. 2006 = 2(3) + 5(400) and then 2007 = 2 (3^4) + 5(1845/5) and then 2008 = 2 (3^2)+(1990/5). Given that 2005 is a possible answer, 2006, 2007 and 2008 are not 2005 is the largest solution.

Siddharth Iyer - 6 years, 2 months ago
Mohamad Zare
Mar 10, 2016

2,7,12,17,22,.…,2002,2007_ 6,11,16,21,…2001,2006_ 18,23,28,…,2003,2008_ For 2005 : 1_ have 2005/3 Or 2_ have 2000 . so 1 it is not integer ,2 in S it is not any 0,or 5 in the last digit!

Afkar Aulia
Dec 22, 2015

Is there even any reason why 5 cannot be in S? The rule never mentions about what isn't a member of S. Even if it's not derived from a known member of S, why can't I include it just because I want it?

Cs ಠ_ಠ Lee
May 24, 2015

Case work!

Based on the clues, we know that the 2 is going to be important. The other two clues are all the other clues we need. Basically we need an exclusionary system that will take out all of the numbers that we don't want: the numbers in the set S.

Exclude numbers that are: 1) Numbers in form 5n+2 2) Numbers in form 3^n * 2 3) Numbers in form 3^n(5n+2) 4) Numbers in form 3^n*2 + 5n

1) It's simple. No numbers that end with 2 or 7 (since multiples of 5 repeat, so 5+2 or 10+2 = 7 or 12 so the units digit shouldn't be 7 or 2.)

2) No doubles of a power of 3. Seems easy enough to classify.

3) No multiples of three that end with 1 or 6, since (5n+2) ends with 2 or 7, and multiplying that by 3 we get 1 or 6.

4) No numbers that end in 6 8 4 or 2. (It's pretty obvi why)

So we test our luck a little here, it can't be numbers that end in 1,2,3,4,6,7,8. Testing the first number that doesn't have those digits, 2005, we get 2005.

Sorry for the bashy solution, I'm bad at math.

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...