Oscar Cunningham<p><a href="https://mathstodon.xyz/tags/Mathober" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Mathober</span></a> <a href="https://mathstodon.xyz/tags/Mathober2024" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Mathober2024</span></a></p><p>The prompt for day 5 was 'Integer Partitions'. In number theory, a partition is a way to write an integer as a sum of positive integers. For example, there are 5 partitions of 4, given by 1+1+1+1, 2+1+1, 2+2, 3+1 and 4. The partition function is the name given to the function that counts how many partitions a given integer has. So the above examples show that p(4)=5. The first few values of the partition function are 1, 1, 2, 3, 5, 7, 11, ....</p><p>In 1937, Hans Rademacher found a complicated formula for the partition function in the form of an infinite series. You can see the full formula here <a href="https://en.wikipedia.org/wiki/Partition_function_(number_theory)#Approximation_formulas" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">en.wikipedia.org/wiki/Partitio</span><span class="invisible">n_function_(number_theory)#Approximation_formulas</span></a>. One interesting feature of this formula is that it allows you to calculate a value of p(x) even when x is not an integer. This was explored and graphed a bit by Fredrik Johansson over at <a href="https://mathoverflow.net/questions/366733/does-rademachers-convergent-series-for-pn-define-an-analytic-function/366805#366805" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">mathoverflow.net/questions/366</span><span class="invisible">733/does-rademachers-convergent-series-for-pn-define-an-analytic-function/366805#366805</span></a>. He points out that when x is n+1/2 for natural n, the infinite series is zero for every term except the first. This then gives you a closed form expression, which he doesn't actually write out because it's awful:</p><p>p(x) = (√(2/3)cosh(π√(2/3)√(x-1/24)) - sinh(π√(2/3)√(x-1/24))/(π√(x-1/24)))/(2√2(x-1/24))</p><p>It's interesting that p(n+1/2) has this closed form formula, because no such formula is known for p(n) itself.</p><p>Of course it would be quite irresponsible to say 'There are 0.8458... ways to write 1/2 as the sum of natural numbers', so I won't.</p><p><a href="https://mathstodon.xyz/tags/Math" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Math</span></a> <a href="https://mathstodon.xyz/tags/Maths" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Maths</span></a> <a href="https://mathstodon.xyz/tags/Mathematics" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Mathematics</span></a> <a href="https://mathstodon.xyz/tags/NumberTheory" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>NumberTheory</span></a> <a href="https://mathstodon.xyz/tags/Combinatorics" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Combinatorics</span></a></p>