Naive permanent vs. Ryser's algorithm in Rust

This example computes the permanent of a square integer matrix in two ways:

• naive_permanent directly enumerates all column permutations.
• ryser_permanent uses Ryser's formula with Gray-code subset traversal.

Algorithm	Cases examined	Time	Extra space
Naive	n! permutations	O(n * n!)	O(n)
Ryser	2^n subsets	O(n * 2^n)	O(n)

Both are exponential, but 2^n grows much more slowly than n!. For example:

Matrix size	Naive permutations	Ryser subsets
4	24	16
8	40,320	256
12	479,001,600	4,096
20	2,432,902,008,176,640,000	1,048,576

Run the example:

cargo run

The program compares both implementations on an 8 x 8 all-ones matrix. Its permanent is 8! = 40,320. Example output (timings vary by machine and build):

Naive: result = 40320, time = 9.1ms
Ryser: result = 40320, time = 84.2us
Naive examines 8! = 40,320 permutations.
Ryser examines 2^8 - 1 = 255 non-empty subsets.

Run the tests:

cargo test

Use cargo run --release for a meaningful timing comparison. The test suite also checks that both algorithms return the same result for several matrices.