Crate approx_int

This library approximates large numbers by calculating the number of bits needed to store a number. The algorithm determines the maximum value that can be represented using this bit length, and then finds the nearest percentage value that can approximately match the original number. This approximation reduces the size of the number while retaining enough information for practical use.

The compact representation of numbers uses a tuple (u8, u8, bool):

• u8 for the number of bits required to store the value.
• u8 for the percentage that describes the degree of approximation.
• bool for storing the sign of the number (true if negative).

In total, this representation uses 24 bits, but you can omit the bool if you are sure the number is positive, then only 16 bits will be needed. The approximate number will generally be smaller than the original, although there are exceptions, especially when working with negative values. For positive numbers, the approximation usually results in a slightly smaller value, but with negative numbers, the approximation could be either smaller or larger than the original.

Structs

SmallValue

The structure stores the resulting number in a compressed format from which an approximated number can be obtained

Traits

SpecialBytes

The trait is used to determine the number of bits required to store a number