FixedPoint64

Implements fixed point numbers in Move language using the Q number format.

Use u128 as underlying data storage. High 64 bits for integer part, Low 64 bits for fractional part.

Similar one used in Uniswap v2 core.

Compile

aptos move compile

Test

aptos move test

Usage

• encode(x: u64): FixedPoint64: Convert u64 integer to FP64

• decode(fp: FixedPoint64): u64: Convert FP64 to u64 integer by rounding down (truncating the fractional part).

  • Examples: 2.00 -> 2, 2.01 -> 2, 2.99 -> 2
  • This is the default way to convert to integer.

• decode_round_up(fp: FixedPoint64): u64: Convert FP64 to u64 integer by rounding up.

  • Examples: 2.00 -> 2, 2.01 -> 3, 2.99 -> 3
  • Be careful when using this. You should not mint more coins than the receipient entitles

• to_u128(fp: FixedPoint64): u128: Get the raw u128 value of FP64

  • The result is NOT the actual integer value

• from_u128(v: u128): FixedPoint64: Create FP64 from raw u128 value

  • The input is NOT the actual integer value

Add as dependency

Add to Move.toml:

[dependencies.FixedPoint64]
git = "https://github.com/ThalaLabs/fixed_point64.git"
rev = "<commit hash>"

And then use in code:

use fixed_point64::fixed_point64;
...
let number = fixed_point64::encode(10);

LICENSE

MIT.