Semi-Squares

I define a "semi-square n n " as a positive integer—which is not a square number—such that there exists a prime p p satisfying n p = x 2 \frac{n}{p}=x^2 for some integer x . x.

How many semi-squares are there between 1 1 and 1000 1000 inclusive?


The answer is 337.

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

Jordan Cahn
Jan 21, 2019

We are looking for integers of the form n = m 2 p n=m^2p , where p p is prime. If n 1000 n\leq 1000 then m 2 1000 p m^2\leq\frac{1000}{p} and m 1000 p m\leq \sqrt{\frac{1000}{p}} . Thus we compute i = 1 π ( 1000 ) 1000 p i = i = 1 168 1000 p i \sum_{i=1}^{\pi(1000)}\left\lfloor\sqrt{\frac{1000}{p_i}}\right\rfloor=\sum_{i=1}^{168}\left\lfloor\sqrt{\frac{1000}{p_i}}\right\rfloor where π ( 1000 ) = 168 \pi(1000)=168 is the prime-counting function (i.e. the number of primes less than 1000 1000 ) and p i p_i denotes the i i th prime number.

Evaluating the function computationally yields 337 337 . I welcome any insight into a simple way to evaluate the sum that can be done by hand.

thanks for the solution. This is a nice first idea

A Former Brilliant Member - 2 years, 4 months ago

I did something similar, but I ran through the squares instead of the primes, so my formula would have been n = 1 500 π ( 1000 n 2 ) \displaystyle\sum_{n=1}^{\lfloor \sqrt{500} \rfloor} \pi\left(\left\lfloor \frac{1000}{n^2} \right\rfloor\right)

I like yours better, just because of how many times I had to call pi(n) in wolfram alpha while adding mine.

Jason Carrier - 2 years, 4 months ago

Just an observation...if we define s ( n ) s(n) to be the number of semi-squares less than or equal to n n , we have

s ( 1000 ) = 2 π ( 1000 ) + 1 s(1000) = 2\pi(1000)+1

This caught my eye, and as I had used a programme to find the answer, it was easy to check that

s ( 10000 ) = 2 π ( 10000 ) + 1 s(10000) = 2\pi(10000)+1 as well. Remarkably, it turns out this is a coincidence, and not in fact generally true; but it does seem to be true that s ( n ) s(n) is very close to 2 π ( n ) 2\pi(n) . I'm sure there's an excellent reason for this, but it eludes me at the moment - does anyone have any ideas why this might be?

Chris Lewis - 2 years, 4 months ago

Log in to reply

Interesting! For p > n 2 p>\frac{n}{2} , there is precisely one semi-square less than n n with p p as a factor... p p itself. The bulk of the semi-squares come from small primes.

Jordan Cahn - 2 years, 4 months ago
X X
Feb 7, 2019
Peter Macgregor
Jan 28, 2019

In mathematica the one liner

Total[Table[PrimePi[1000/n^2],{n,1,31}]]

does the trick!

The idea is to take each of the squares and multiply them by prime numbers to produce semi-squares. Each entry in the table counts the number of semi-squares less than 1000 coming from each square.

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...