Find all triples of positive integers that satisfy
Prove that your list is complete (i.e., show there are no other positive integer solutions).
Find all triples of positive integers that satisfy
Prove that your list is complete (i.e., show there are no other positive integer solutions).
In order to solve this problem, we will first analyse possible sets of three positive integers that will satisfy the equation:
Undertaking a simple analysis of the problem shows us a number of solutions that satisfy these criteria. We will denote each individual solution S_n as S_n = [a,b,c]
We will also denote that a \leq b \leq c in each case, as it simplifies the need to rearrange the answers without changing the values.
We can see logically that adding three thirds together will give us 1/3 + 1/3 +1/3 = 1.
Cross-verification:
Adding one half to one quarter, and then adding another quarter will also give us an answer of 1, and therefore we can see that 1/2 + 1/4 + 1/4 = 1.
Cross-verification:
Adding 1/2 to 1/3 gives us \frac{5}{6}. Adding 1/6 here will give us 1.
Cross-verification:
There are no other solutions that satisfy the equation. So, how do we prove this empirically?
We will maintain our previous logic of a \leq b \leq c, as once again, it makes no difference to the possible outcomes.
Firstly, it is important to denote that \{a,b,c\} \in \mathbb{Z}^{+}, as per the question requirements. Since 0 is not a positive integer, it will not be included in any valid solution. Inclusion of 0 would also require us to calculate \frac{1}{0}, which would also be impossible.
Without 0, it becomes impossible to include 1, as the only way to reach the exact value \frac{1}{1} for our answer, where one of the values is 1, would be for a=0, b=1 and c=1 to be used, which is impossible as previously stated.
Our test is to verify whether \frac{1}{a}+\frac{1}{b}+\frac{1}{c} = 1. Let it also be noted that in all cases, as the value of x increases, the value of \frac{1}{x} will always decrease. Therefore, for any answer where \frac{1}{a}+\frac{1}{b}+\frac{1}{c} \lt 1, increasing any of a, b, or c would not make a difference, as an increase in the denominator values would always lead to the final value trending further away from 1.
Keeping our a \leq b \leq c logic in mind, if we analyse potential solutions in ascending order such as [4,4,4] being followed by [4,4,5] and then [4,4,6], etc., a proof that [4,4,4] fails our test would also empirically prove that all higher triples, for instance [4,4,5] or [4,4,6], or [5,6,7], as examples, would fail too.
To show that [4,4,4] fails, we can simply input the values:
therefore, it is impossible for a valid solution to contain an a value that is greater than or equal to 4.
Based on the elimination constraints set out above, it becomes apparent that our only plausible a values are either a=2 or a=3, on the basis that we have eliminated a=1 and any a \geq 4, and \{a \in \mathbb{Z}^+\}. We can now take a look at both cases and determine further elimination steps for b and c.
An evaluation of
where a=2, gives us
In other words, the product of b and c must be equal to double the sum of b and c.
The only positive integer pair that would satisfy this equation in full, is b = 4 and c=4.
To prove that [4,4] is the only positive integer pair where the product is equal to twice the sum, we can assess the relationship xy = 2(x+y), \{x,y \in \mathbb{Z}^+\}, which should only prove true when x and y are both equal to 4.
Since there is no inverse relationship here, and we are dealing only with positive integers, increasing the value of x or y (and subsequently xy) will always lead to a respective increase in the overall value of 2(x+y). Similarly, reducing the value of x or y (and subsequently xy) will always reduce the overall value of 2(x+y).
Let us not forget that the values of xy and 2(x+y) are equivalent when x and y both equal 4. We can now assess [4,4], and the pairs that neighbour it spacewise (excluding repeats) i.e [3,4] to the left and [4,5] to the right.
For [3,4], the 2(x+y) value of 14 is greater than the xy value of 7.
It can therefore be empirically stated that any pairs with a value below [3,4], (such as [3,3], [2,3] and so on) will follow the same pattern, and none will be equal.Similarly, for [4,5], the 2(x+y) value of 18 is less than the xy value of 20. It can therefore be empirically stated that any pairs with a value above [4,5], (such as [5,5], [5,6] and so on) will follow the same pattern, and none will be equal.
Applying both of these logical steps together proves undeniably that [4,4] is the only possible integer pair \in \mathbb{Z}^+ that maintains an equivalence between bc and 2(b+c).
Hence, when a = 2, b must equal 4, and c must equal 4. This means that the only possible solution when a=2 is one where [a,b,c] = [2,4,4].
[/details]
Taking a look at our original equation, \frac{1}{a}+\frac{1}{b}+\frac{1}{c}=1, we can see that
Substituting 3 for the a-value here gives us
In other words, we now need to show that [3,3] is the only [b,c] pair where twice the product is equivalent to three times the sum.
Taking a look at spacewise neighbour [2,3], we can see that twice the product, 2(2 \times 3) = 12 is less than three times the sum, 3(2+3)= 15.
Taking a look at spacewise neighbour [3,4], we can see that twice the product, 2(3 \times 4) = 24 is greater than three times the sum, 3(3+4)= 21.
Similar to before, there are no negative values or inverse relationships in our set. Therefore, it can be stated that [3,3] is the only pair that meets our criteria.
This proves empirically that when a = 3, b and c must also equal 3 for a valid solution.
As shown, there are only three solutions for a, b and c that satisfy the equation
These are the solutions, denoted [a,b,c] where the values are