Calculate the integer square root of a nonnegative number n, that is, the largest integer r with r*r <= n. Throws an error on negative input.

Test whether the argument is a square. After a number is found to be positive, first isPossibleSquare is checked, if it is, the integer square root is calculated.

Returns Nothing if the argument is not a square, Just r if r*r == n and r >= 0. Avoids the expensive calculation of the square root if n is recognized as a non-square before, prevents repeated calculation of the square root if only the roots of perfect squares are needed. Checks for negativity and isPossibleSquare.

Calculate the integer cube root of an integer n, that is the largest integer r such that r^3 <= n. Note that this is not symmetric about 0, for example integerCubeRoot (-2) = (-2) while integerCubeRoot 2 = 1.

Test whether an integer is a cube.

Returns Nothing if the argument is not a cube, Just r if n == r^3.

Calculate the integer fourth root of a nonnegative number, that is, the largest integer r with r^4 <= n. Throws an error on negaitve input.

Test whether an integer is a fourth power. First nonnegativity is checked, then the unchecked test is called.

Returns Nothing if n is not a fourth power, Just r if n == r^4 and r >= 0.

Calculate an integer root, integerRoot k n computes the (floor of) the k-th root of n, where k must be positive. r = integerRoot k n means r^k <= n < (r+1)^k if that is possible at all. It is impossible if k is even and n < 0, since then r^k >= 0 for all r, then, and if k <= 0, integerRoot raises an error. For k < 5, a specialised version is called which should be more efficient than the general algorithm. However, it is not guaranteed that the rewrite rules for those fire, so if k is known in advance, it is safer to directly call the specialised versions.

isKthPower k n checks whether n is a k-th power.

exactRoot k n returns Nothing if n is not a k-th power, Just r if n == r^k. If k is divisible by 4, 3 or 2, a residue test is performed to avoid the expensive calculation if it can thus be determined that n is not a k-th power.

isPerfectPower n checks whether n == r^k for some k > 1.

highestPower n produces the pair (b,k) with the largest exponent k such that n == b^k, except for abs n <= 1, in which case arbitrarily large exponents exist, and by an arbitrary decision (n,3) is returned.

First, by trial division with small primes, the range of possible exponents is reduced (if p^e exactly divides n, then k must be a divisor of e, if several small primes divide n, k must divide the greatest common divisor of their exponents, which mostly will be 1, generally small; if none of the small primes divides n, the range of possible exponents is reduced since the base is necessarily large), if that has not yet determined the result, the remaining factor is examined by trying the divisors of the gcd of the prime exponents if some have been found, otherwise by trying prime exponents recursively.

powMod b e m computes (b^e) `mod` m in effective way. An exponent e must be non-negative, a modulo m must be positive. Otherwise the behaviour of powMod is undefined.

>>> powMod 2 3 5
3
>>> powMod 3 12345678901234567890 1001
1

See also powMod and powSomeMod.

For finite numeric types (Int, Word, etc.) modulo m should be such that m^2 does not overflow, otherwise the behaviour is undefined. If you need both to fit into machine word and to handle large moduli, take a look at powModInt and powModWord.

>>> powMod 3 101 (2^60-1 :: Integer)
1018105167100379328 -- correct
>>> powMod 3 101 (2^60-1 :: Int)
1115647832265427613 -- incorrect due to overflow
>>> powModInt 3 101 (2^60-1 :: Int)
1018105167100379328 -- correct