Variable EQUALITY_OPTIONSConst

EQUALITY_OPTIONS: {
    DOUBLE: {
        precision: number;
        unit: number;
    };
    DOUBLE_HIGH: {
        precision: number;
        unit: number;
    };
    DOUBLE_LOW: {
        precision: number;
        unit: number;
    };
    FULL_DOUBLE: {
        precision: number;
        unit: number;
    };
    FULL_HALF: {
        precision: number;
        unit: number;
    };
    FULL_SINGLE: {
        precision: number;
        unit: number;
    };
    HALF: {
        precision: number;
        unit: number;
    };
    SINGLE: {
        precision: number;
        unit: number;
    };
} = ...

Collection of equality options.

Type declaration

  • DOUBLE: {
        precision: number;
        unit: number;
    }

    Default equality options.
    Handles precision of the f64 datatype with 10 binary places of leaway.
    Precision is 43 and unit is 1.

    • precision: number
    • unit: number
  • DOUBLE_HIGH: {
        precision: number;
        unit: number;
    }

    Standard equality options for the f64 datatype, with a high unit.
    The unit is 2e-12, thus anything smaller than about 2e-24 will be considered zero.

    • precision: number
    • unit: number
  • DOUBLE_LOW: {
        precision: number;
        unit: number;
    }

    Standard equality options for the f64 datatype, with a small unit.
    The unit is 2e-24, thus anything smaller than about 2e-72 will be considered zero.

    • precision: number
    • unit: number
  • FULL_DOUBLE: {
        precision: number;
        unit: number;
    }

    The full equality of the f64 datatype.
    Has no leaway.
    WARNING: This does not account for any floating point errors.

    • precision: number
    • unit: number
  • FULL_HALF: {
        precision: number;
        unit: number;
    }

    The full equality of the f16 datatype.

    • precision: number
    • unit: number
  • FULL_SINGLE: {
        precision: number;
        unit: number;
    }

    The full equality of the f32 datatype.

    • precision: number
    • unit: number
  • HALF: {
        precision: number;
        unit: number;
    }

    Lowest precision equality options.
    Handles precision of the f16 datatype with 3 binary places of leaway.

    • precision: number
    • unit: number
  • SINGLE: {
        precision: number;
        unit: number;
    }

    Lower precision equality options.
    Handles precision of the f32 datatype with 4 binary places of leaway.

    • precision: number
    • unit: number

Settings

Member Visibility

Theme

Generated using TypeDoc