Num represents a signed 128-bit integer in two's complement.
Calculations wrap around and overflow is ignored.
For a discussion of the algorithms, look at Knuth's volume 2,
Semi-numerical Algorithms section 4.3.1.
Adapted from the Apache ORC C++ implementation Abs returns a new decimal128.Num that contains the absolute value of n( Num) Add(rhs Num) Num while the code would be simpler to just do lsh/rsh and add
it turns out from benchmarking that calling SetBits passing
in the words and negating ends up being >2x faster Cmp compares the numbers represented by n and other and returns:
+1 if n > other
0 if n == other
-1 if n < other( Num) Div(rhs Num) (res, rem Num) FitsInPrecision returns true or false if the value currently held by
n would fit within precision (0 < prec <= 38) without losing any data. Greater returns true if the value represented by n is > other GreaterEqual returns true if the value represented by n is >= other HighBits returns the high bits of the two's complement representation of the number. IncreaseScaleBy returns a new decimal128.Num with the value scaled up by
the desired amount. Must be 0 <= increase <= 38. Any data loss from scaling
is ignored. If you wish to prevent data loss, use Rescale which will
return an error if data loss is detected. Less returns true if the value represented by n is < other LessEqual returns true if the value represented by n is <= other LowBits returns the low bits of the two's complement representation of the number.( Num) Mul(rhs Num) Num Negate returns a copy of this Decimal128 value but with the sign negated( Num) Pow(rhs Num) Num ReduceScaleBy returns a new decimal128.Num with the value scaled down by
the desired amount and, if 'round' is true, the value will be rounded
accordingly. Assumes 0 <= reduce <= 38. Any data loss from scaling
is ignored. If you wish to prevent data loss, use Rescale which will
return an error if data loss is detected. Rescale returns a new decimal128.Num with the value updated assuming
the current value is scaled to originalScale with the new value scaled
to newScale. If rescaling this way would cause data loss, an error is
returned instead. Sign returns:
-1 if x < 0
0 if x == 0
+1 if x > 0( Num) Sub(rhs Num) Num( Num) ToBigFloat(scale int32) *big.Float ToFloat32 returns a float32 value representative of this decimal128.Num,
but with the given scale. ToFloat64 returns a float64 value representative of this decimal128.Num,
but with the given scale.( Num) ToString(scale int32) string
Num : github.com/apache/arrow-go/v18/arrow/decimal.Num[decimal.Decimal128]
func FromBigInt(v *big.Int) (n Num)
func FromFloat32(v float32, prec, scale int32) (Num, error)
func FromFloat64(v float64, prec, scale int32) (Num, error)
func FromI64(v int64) Num
func FromString(v string, prec, scale int32) (n Num, err error)
func FromU64(v uint64) Num
func GetHalfScaleMultiplier(pow int) Num
func GetMaxValue(prec int32) Num
func GetScaleMultiplier(pow int) Num
func Max(first Num, rest ...Num) Num
func Min(first Num, rest ...Num) Num
func New(hi int64, lo uint64) Num
func Num.Abs() Num
func Num.Add(rhs Num) Num
func Num.Div(rhs Num) (res, rem Num)
func Num.IncreaseScaleBy(increase int32) Num
func Num.Mul(rhs Num) Num
func Num.Negate() Num
func Num.Pow(rhs Num) Num
func Num.ReduceScaleBy(reduce int32, round bool) Num
func Num.Rescale(originalScale, newScale int32) (out Num, err error)
func Num.Sub(rhs Num) Num
func Max(first Num, rest ...Num) Num
func Max(first Num, rest ...Num) Num
func Min(first Num, rest ...Num) Num
func Min(first Num, rest ...Num) Num
func Num.Add(rhs Num) Num
func Num.Cmp(other Num) int
func Num.Div(rhs Num) (res, rem Num)
func Num.Greater(other Num) bool
func Num.GreaterEqual(other Num) bool
func Num.Less(other Num) bool
func Num.LessEqual(other Num) bool
func Num.Mul(rhs Num) Num
func Num.Pow(rhs Num) Num
func Num.Sub(rhs Num) Num
func github.com/apache/arrow-go/v18/arrow/decimal256.FromDecimal128(n Num) decimal256.Num
func github.com/apache/arrow-go/v18/arrow/scalar.NewDecimal128Scalar(val Num, typ arrow.DataType) *scalar.Decimal128
var MaxDecimal128
Package-Level Functions (total 12)
FromBigInt will convert a big.Int to a Num, if the value in v has a
BitLen > 128, this will panic.
FromFloat32 returns a new decimal128.Num constructed from the given float32
value using the provided precision and scale. Will return an error if the
value cannot be accurately represented with the desired precision and scale.
FromFloat64 returns a new decimal128.Num constructed from the given float64
value using the provided precision and scale. Will return an error if the
value cannot be accurately represented with the desired precision and scale.
FromI64 returns a new signed 128-bit integer value from the provided int64 one.
The pages are generated with Goldsv0.8.2. (GOOS=linux GOARCH=amd64)
Golds is a Go 101 project developed by Tapir Liu.
PR and bug reports are welcome and can be submitted to the issue list.
Please follow @zigo_101 (reachable from the left QR code) to get the latest news of Golds.
