Beware, Just Integers!

Algebra Level 5

Consider a geometric progression with distinct numbers which belong to the set { 100 , 101 , 102 , , 999 , 1000 } {\{100,101,102,\ldots,999,1000}\} .

What is the maximum number of terms it can have?

There's no such maximum terms 5 7 4 6

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.

3 solutions

Nihar Mahajan
Sep 13, 2015

Note that we want maximum terms. Suppose the common ratio is 2 2 . Then the GP with maximum term is the one whose first term is smallest i.e 100 100 . So the GP is 100 , 200 , 400 , 800 100,200,400,800 . But this is not the GP with maximum terms.Since the terms are distinct the common ratio is not equal to 1 1 . So we conclude that if r r is common ratio , to get maximum terms in GP , optimally , the value of r r must be bound : 1 < r < 2 1 < r < 2 .

So let r = 3 2 r=\dfrac{3}{2} . For the terms of GP to be an integer and to have maximum terms , the first term must optimally be as small as possible and can be represented as 2 n 2^n where n n must be maximum as possible. And we have 2 7 = 128 2^7 = 128 . Thus the GP required is 128 , 192 , 288 , 432 , 648 , 972 128,192,288,432,648,972 with 6 \boxed{6} terms.

Nice solution, Nihar. This is indeed the only GP under the given conditions that has 6 6 terms. To have 7 7 terms we would require the least integer of the sequence to be divisible by 2 6 , 2^{6}, and 128 128 is the least such integer. But as you've shown, with r = 3 / 2 r = 3/2 we can only get 6 6 terms under 1000. 1000.

To have 6 6 terms we require the least integer in the sequence to be divisible by p 5 p^{5} for some prime p . p. You've already dealt with p = 2. p = 2. With p = 3 p = 3 we would then look at 3 5 = 243 3^{5} = 243 and r = 4 / 3 , r = 4/3, which would yield the sequence 243 , 324 , 432 , 576 , 768 , 1024 , 243, 324, 432, 576, 768, 1024, which just misses. So the sequence you've found is uniquely the maximum in length. (Those first three terms are interesting in that they are each a permutation of (2,3,4). This looks like the start of an interesting question ....)

A follow-up question could be to find the number of GPs under the given conditions (and with maximized length) that have precisely 5 terms. I get a total of 5; for r = 3 / 2 r = 3/2 the first terms would be 112 , 144 , 160 , 176 , 112, 144, 160,176, and for r = 4 / 3 r = 4/3 the first term would be 162. 162.

Brian Charlesworth - 5 years, 9 months ago

Log in to reply

Yeah! great!

Nihar Mahajan - 5 years, 9 months ago

For your follow up question......I am getting 9 such GPs. If r=3/2 first term can be (16)k where k can be 7,8,9,10,11,12. If r=4/3 then first term can be (81)k where k can be 2 and 3. If r=5/4 then first term will be (256)k where k can be 1. Am I doing something wrong??? @Brian Charlesworth

Ghanshyam Dhoot - 3 years, 7 months ago

Log in to reply

You're mostly right; I did miss a few. We wouldn't include 16 × 8 16 \times 8 since there would be a 6th term 972 972 , and we want precisely 5 terms, but otherwise your other 8 GPs will work.

Brian Charlesworth - 3 years, 7 months ago

Same way. Nice solution.

Kushagra Sahni - 5 years, 9 months ago

The value of r is not bound as you stated as 1>r>2 rather it should be 1<r<2

Devendra Kumar Singh - 5 years, 8 months ago

Log in to reply

Oh , thanks , that was really bad typing mistake ;P

Nihar Mahajan - 5 years, 8 months ago
Andrea Palma
Sep 19, 2015

Lets's try to assume that there is a GP with at least 7 terms. This means that we have a starting number a a and a ratio r r such that

a , a r , a r 2 , , a r 6 { 100 , 101 , , 1000 } N { 0 } a, ar, ar^2, \ldots , ar^6 \in \left\{100, 101, \ldots , 1000 \right\} \subset \mathbb N \setminus \left\{ 0 \right\}

Now, since all the terms must be integers this means that a N a \in \mathbb N and a r N ar \in \mathbb N . This implies that r Q r \in \mathbb Q in fact if r r is irrational, and since a a is natural, we would have a r ar irrational. Morever r r is positive otherwise a r ar would be not positive as it must be since it belongs to { 100 , , 1000 } \left\{ 100, \ldots, 1000 \right\} . So there are two positive integers m , d N m,d \in \mathbb N such that

r = m d r = \dfrac{m}{d}

Note that m d m \not = d otherwise r = 1 r=1 and all the ters of the GP would be the same (they must be distinct).

Since a r 6 = a m 6 d 6 N ar^6 = a \dfrac{m^6}{d^6}\in \ \mathbb N this implies d 6 a d^6 \vert a . How many d d are such that d 6 d^6 divides the number a a belonging to { 100 , , 1000 } \left\{ 100, \ldots, 1000 \right\} ? We have few choices for d d .

First choiche d = 1 d=1 . That implies a a can be whatever integer number from 100 100 to 1000 1000 . But we must also have a m 6 { 100 , , 1000 } am^6 \in \left\{ 100, \ldots, 1000 \right\} and must be m 2 m \geq 2 , and a 100 a \geq 100 . Now a m 6 100 2 6 = 6400 am^6 \geq 100 \cdot 2^6 = 6400 and it is not in { 100 , , 1000 } \left\{ 100, \ldots, 1000 \right\} . This means d d cannot be 1 1 .

Second choiche d = 2 d=2 . That means 64 64 divides a a . So that a a can be one of these values 128 = 64 2 , 192 = 64 3 , , 960 = 64 15 128 = 64 \cdot 2, 192 = 64 \cdot 3, \ldots , 960 = 64 \cdot 15 so that we have a = 64 x a = 64 \cdot x with x { 2 , 3 , , , 15 } x \in \left\{ 2,3,, \ldots , 15 \right\} .What about m m ? m m could be 1 1 but this would led to a r 6 = a 1 64 = x { 100 , , 1000 } a r^6 = a \dfrac{1}{64} = x \in \left\{ 100, \ldots, 1000 \right\} while x 15 x \leq 15 . So m 1 m \not = 1 . m 3 m \geq 3 but this would let to a r 6 = x m 6 2 729 = 1458 { 100 , , 1000 } ar^6 = xm^6 \geq 2\cdot 729 = 1458 \in \left\{ 100, \ldots, 1000 \right\} that is not true. This means no vlues for m m can suit, and definitely d d cannot be 2.

Third choice d = 3 d = 3 . This means 729 729 divides a a that leaves only one possibility a = 729 a = 729 . The fact that a r 6 { 100 , , 1000 } ar^6 \in \left\{ 100, \ldots, 1000 \right\} transaltes in m 6 { 100 , , 1000 } m^6 \in \left\{ 100, \ldots, 1000 \right\} but the only integer whose sixth power belongs in that range is 3 3 and m m must be distinct from d = 3 d=3 . So there are no choices for m m and thid means d d cannot be 3 3 .

Fourth choice d 4 d \geq 4 . This leads immediately to a contraddiction because it would imply that d 6 4096 d^6 \geq 4096 is a divisor of a 1000 a \leq 1000 . So it can't be d 4 d \geq 4 .

Every choiche for d d takes to a contraddiction, so we proved that there are no GP with at least 7 terms in { 100 , , 1000 } \left\{ 100, \ldots, 1000 \right\} .

Otherwise it is easy to find a GP with 6 terms (the same argument from the previous proof can be used). Namely the one we get with a = 128 a = 128 and r = 3 2 r = \dfrac{3}{2}

128 , 192 , 288 , 432 , 648 , 972. 128, \ 192, \ 288, \ 432, \ 648, \ 972.

So we can conclude that the maximum number of terms the GP can have is 6.

Lu Chee Ket
Dec 17, 2015

Result of search:

1
2
3
4
5
6
128
192
288
432
648
972

Only one set is available with a = 128 and r = 1.5 for maximum of 6 terms. There are six sets for 5 terms and zero set for 7 terms. Search fineness was 0.0001 to 3 for r, with a = 100 to 999. Finest set was found for 5 terms of a = 256 and r = 1.25 which is {256, 320, 400, 500, 625}.

Answer: 6 \boxed{6}

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...