KahanSum
A simple quadruple precision sum (around 108 bit) based on Kahan's summation method. This is used in Stats, to maintain higher precision sums for statistical moment computation.
Constructors
| Constructor | Description |
new(sum, carry)
Signature: (sum:float * carry:float) -> unit
|
|
new(sum)
Signature: sum:float -> unit
|
Instance members
| Instance member | Description |
Add(value)
Signature: value:float -> unit
|
|
Add(sum)
Signature: sum:KahanSum -> unit
|
|
Sub(value)
Signature: value:float -> unit
|
|
Sub(sum)
Signature: sum:KahanSum -> unit
|
|
Value
Signature: float
|
Static members
| Static member | Description |
op_Addition(sum, value)
Signature: (sum:KahanSum * value:float) -> KahanSum
|
|
op_Addition(sum0, sum1)
Signature: (sum0:KahanSum * sum1:KahanSum) -> KahanSum
|
|
op_Subtraction(sum, value)
Signature: (sum:KahanSum * value:float) -> KahanSum
|
|
op_Subtraction(sum0, sum1)
Signature: (sum0:KahanSum * sum1:KahanSum) -> KahanSum
|
