20 Feb
2014
20 Feb
'14
4:28 p.m.
An alternative repair for #37 is that the LHS is left alone, with the right-hand side changed from p(n) to q(n), and the description changed to partitions with _unequal_ terms. Quoting Dan Asimov <dasimov@earthlink.net>:
Right, Mike nailed it: 1 + x^k should be 1/(1-x^k). Duh.
On Feb 20, 2014, at 2:35 PM, Dan Asimov <dasimov@earthlink.net> wrote:
Should #37 have a minus sign, not a plus sign, on the LHS?
(And I had to look up why #41 is wrong. Why is it merely ?wrong? ?) The error term is given as o(n); correct is O(1/n), but o(n) is also true, though not particularly useful. Rich