MedVision ad

Solving a number theoretical problem (1 Viewer)

mamun11

New Member
Joined
Sep 10, 2014
Messages
9
Gender
Male
HSC
N/A
How many pairs(A,B) are there up to n such that GCD(A,B)=B without those pairs(A,B), where B^2=A.
If we consider 5 as n we have 25 possible pairs.They are (1,1),(1,2),(1,3),(1,4),(1,5),(2,1),(2,2),(2,3),(2,4),(2,5),...,(5,5).

Of them, 8 pairs satisfy the above condition.My main objective is to find out the number
of such pairs if n is given with a efficient method.One of my friends showed me a method but he didn't state the reasoning behind it.

Here,I'm presenting his method for n=10.

2 * ((floor(10/1) - 1) + (floor(10/2) - 2) + (floor(10/3) - 3))

= 2 * ((10 - 1) + (5 - 2) + (3 - 3))

= 2 * (9 + 3 + 0)

= 24

Can anyone explain the reasoning behind the above method in details and clearly?
 
Last edited:

seanieg89

Well-Known Member
Joined
Aug 8, 2006
Messages
2,662
Gender
Male
HSC
2007
Okay, well first note that gcd(A,B)=B is equivalent to just saying that B divides A. (Let me know if you don't see why this is the case).

So let's count the number of pairs of positive integers not exceeding N such that B|A.

This is the same as letting B vary between 1 and N, and counting the number of multiples of B that do not exceed N.
'
This is just

Given the non-square condition, we need to subtract 1 when B does not exceed

Hence, the number you are looking for is

 

seanieg89

Well-Known Member
Joined
Aug 8, 2006
Messages
2,662
Gender
Male
HSC
2007
As an extension, you might find it interesting to try and prove that



where is the quantity in the question and log is the natural log.
 

Users Who Are Viewing This Thread (Users: 0, Guests: 1)

Top