We know that any Diophantine equation of the form $ax + by = c$ has either no solutions, or infinite solutions of the form:

$$x = x_0 + n\frac{b}{(a, b)}$$ $$y = y_0 - n\frac{a}{(a, b)}$$

Where $n$ is any integer, $(x_0, y_0)$ is one solution, and $(a, b)=\gcd(a, b)$.

I use this to attempt to answer the question to the best of my abilities, but I get stuck at one part:

By inspection, we can see that $8(2) + 5(-3) = 1$, and so $8(2c) + 5(-3c) = c$. Using the above equations, we find that:

$$x=2c + 5n$$ $$y=-3c - 8n$$

Knowing that both of these values will be above zero, we can set:

$$x >0$$ $$2c + 5n >0$$ $$n > \frac{-2c}{5}$$


$$y>0$$ $$-3c-8n>0$$ $$n<\frac{-3c}{8}$$

Thus, $\frac{-2c}{5}<n<\frac{-3c}{8}$ will yield positive solutions. However, we only want one positive solution. Here's where I get stuck.

Am I able to simply say $\frac{-2c}{5}+1 =\frac{-3c}{8}$? Because it may be that the difference is slightly more than $1$, and so the value of $c$ would differ; $\frac{-2c}{5} $isn't necessarily an integer. Moreover, when I do attempt to say $\frac{-2c}{5}+1 =\frac{-3c}{8}$ anyways, I finish with $-16 < n < -15$, which leaves no answer for $n$ as an integer.

I would appreciate any help.


Your reasoning is correct up to the point where you deal with only one solution. While the condition of $\frac{-2c}{5}+1 = \frac{-3c}{8}$ really does guarantee that there is exactly one integer lying strictly between the lower and upper bound (assuming that $\frac{2c}{5}$ is not an integer), it's far too restrictive. For example, there is one integer between $0.99$ and $1.01$, yet both numbers differ only by $0.02$. On the other end, there is just one integer between $0.01$ and $1.99$, which differ by $1.98$.

Sometimes it's easier to ask for less and get more: Certainly, the difference between upper and lower bound must be smaller or equal to $2$, otherwise there would surely be (at least) two integers in between them. This gives us the inequality $$\frac{-3c}{8}+\frac{2c}{5}\leq 2$$ which simplifies to $c\leq 80$.

Now, here comes the tricky part. If you're interested in the actual list of numbers $c$ satisfying the "only one solution" condition, you might need to go through a bit of tedious work of checking which values of $c$ actually (it turns out to be shorter than it looks at the first glance, but still somewhat laborous).

The count of such numbers $c$ can be determined with considerably less effort thanks to this useful observation (which is not to difficult to prove): For any $1\leq c\leq 40$, exactly one of $c$ and $(c+40)$ has exactly one exactly one solution in positive integers corresponding to it. Thus, there are $40$ values of $c$ satisfying your conditions.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.