speed up gaspari cohn

[mathause]

From #109 (comment): it should be possible to speed up calculating the gaspari cohn function by getting rid of the exponents:

-1 - 5 / 3 * r_s ** 2 + 5 / 8 * r_s ** 3 + 1 / 2 * r_s ** 4 - 1 / 4 * r_s ** 5
+1 - (5 / 3 + (5 / 8 + (1 / 2 - 1 / 4 * r_s) * r_s) * r_s) * r_s**2

admittedly a micro optimisation and making it more difficult to read but about 10 times faster (double check) because exponentiation is expensive.

See also #158

[mathause]

Removing from the milestone again. This remains a not-critical good first issue.