MHB Compute $\sum_{n=1}^{2020} (f(n)-g(n))$

  • Thread starter Thread starter anemone
  • Start date Start date
Click For Summary
The discussion centers on calculating the sum $\sum_{n=1}^{2020} (f(n)-g(n))$, where $f(n)$ represents the sum of all positive integers $b$ that allow integer solutions for the quadratic equation $x^2 + bx + n = 0$, and $g(n)$ represents the sum of all positive integers $c$ for which the linear equation $cx + n = 0$ has integer solutions. The key challenge is to determine the conditions under which these equations yield integer solutions and how to effectively compute the difference between $f(n)$ and $g(n)$. Participants explore the implications of integer solutions on the values of $b$ and $c$, leading to insights on the nature of $f(n)$ and $g(n)$. The final goal is to evaluate the overall sum from 1 to 2020, highlighting the relationship between the two functions. The discussion emphasizes the mathematical properties and calculations necessary to arrive at the solution.
anemone
Gold Member
MHB
POTW Director
Messages
3,851
Reaction score
115
For non-zero integers $n$, let $f(n)$ be the sum of all positive integers $b$ for which all solutions $x$ to $x^2+bx+n=0$ are integers and let $g(n)$ be the sum of all positive integers $c$ for which all solutions $x$ to $cx+n=0$ are integers. Compute $\displaystyle \sum_{n=1}^{2020} (f(n)-g(n))$.
 
Mathematics news on Phys.org
anemone said:
For non-zero integers $n$, let $f(n)$ be the sum of all positive integers $b$ for which all solutions $x$ to $x^2+bx+n=0$ are integers and let $g(n)$ be the sum of all positive integers $c$ for which all solutions $x$ to $cx+n=0$ are integers. Compute $\displaystyle \sum_{n=1}^{2020} (f(n)-g(n))$.
The product of the solutions of $x^2 + bx + n = 0$ is $n$. So if both solutions are integers then they must form a factorisation of $n$. Also, if the solution of $cx+n=0$ is an integer then it must be a factor of $n$. So the ingredients of the sums $f(n)$ and $g(n)$ must all arise from factorisations of $n$.

If $n=st$ is a factorisation of $n$ as a product of two positive integers then the solutions of $x^2 + (s+t)x + n = 0$ are $x= -s$ and $x = -t$. So $b=s+t$ is one of the values of $b$ that go into the sum $f(n)$. Also, $x=-t$ is the solution of $sx + n = 0$ and $x=-s$ is the solution of $tx+n=0$. So $s$ and $t$ are two of the values of $c$ that go into the sum $g(n)$. Therefore the net contribution of the factorisation $n=st$ to $f(n) - g(n)$ is $(s+t) - (s+t) = 0$. Summing over all possible factorisations of $n$, you see that $f(n) - g(n) = 0$.

But that argument goes wrong if $n$ is a perfect square, because the factorisation $n = s^2$ gives a contribution $2s$ to $f(n)$, but it only contributes $s$ to $g(n)$. Therefore $f(s^2) - g(s^2) = s$.

The largest square that is less than $2020$ is $44^2 = 1996$. Therefore $$ \sum_{n=1}^{2020} (f(n)-g(n)) = \sum_{s=1}^{44}s = \tfrac12*44*45 = 990.$$
 
Seemingly by some mathematical coincidence, a hexagon of sides 2,2,7,7, 11, and 11 can be inscribed in a circle of radius 7. The other day I saw a math problem on line, which they said came from a Polish Olympiad, where you compute the length x of the 3rd side which is the same as the radius, so that the sides of length 2,x, and 11 are inscribed on the arc of a semi-circle. The law of cosines applied twice gives the answer for x of exactly 7, but the arithmetic is so complex that the...

Similar threads

Replies
1
Views
1K
Replies
9
Views
2K
Replies
1
Views
1K
Replies
1
Views
1K
Replies
1
Views
1K
Replies
2
Views
1K