When we first came up with two stack rules, we suppose there are n numbers to be added to the stack, which are 1 to n in order.
First point: if integer x is included in the stack (that is, X is the top element of the stack), the integer less than X will be included in the stack.
Second: If the integer y is used to exit the stack (that is, Y is the top element of the stack), the integer greater than y must have exceeded the stack.
The following two rules are available:X is known to have been output from the stack. If y is out of the stack and Y is <X, Z must have been output from the stack for integer Z (z> Y & Z <= X.
Then we can draw another conclusion:If y = 1 is just out of the stack, and Y is the K out of the stack, 2-K has been out of the stack, and K + 1-N has not been in the stack.
After understanding the above reasoning, we should be able to quickly introduce the hdu1023 formula:F [N] = sum (F [I] * f [n-I-1]), where 0 <= I <= n-1;
The formula reasoning is as follows:
For N trains, the numbers are from 1 to n
If the first row after the first row is in the 1st position, 1 first enters the stack, and then immediately exits the stack. The rest are still ordered, with the number being n-1, so the number of outgoing stack order num = f [n-1];
If the first post-exit position is 2nd, the conclusion above indicates that the first post-exit position must be 2, that is, 1 does not move after stack entry, 2 in and out, and 1 in and out, if the number of remaining n-2 is random, num = f [1] * f [N-2];
If the first post-exit position is 3rd, the conclusion above indicates that the first part of the post-exit position must be 2 and 3, that is, 1 does not move after the stack entry, and 2 and 3 are random, 1 out, the number of remaining n-3 is random, then num = f [2] * f [n-3];
............
............
Similarly, if f [0] = 1, a formula can be provided.
Note: hdu1023 has a large result and must be processed by a large integer.