Tensor.Truncate 方法

定义

重载

Truncate<T>(ReadOnlyTensorSpan<T>)

计算指定张量中数字的元素截断。

Truncate<T>(ReadOnlyTensorSpan<T>, TensorSpan<T>)

计算指定张量中数字的元素截断。

Truncate<T>(ReadOnlyTensorSpan<T>)

Source:
TensorExtensions.cs

计算指定张量中数字的元素截断。

public:
generic <typename T>
 where T : System::Numerics::IFloatingPoint<T> static System::Numerics::Tensors::Tensor<T> ^ Truncate(System::Numerics::Tensors::ReadOnlyTensorSpan<T> % x);
public static System.Numerics.Tensors.Tensor<T> Truncate<T> (in System.Numerics.Tensors.ReadOnlyTensorSpan<T> x) where T : System.Numerics.IFloatingPoint<T>;
static member Truncate : ReadOnlyTensorSpan -> System.Numerics.Tensors.Tensor<'T (requires 'T :> System.Numerics.IFloatingPoint<'T>)> (requires 'T :> System.Numerics.IFloatingPoint<'T>)
Public Function Truncate(Of T As IFloatingPoint(Of T)) (ByRef x As ReadOnlyTensorSpan(Of T)) As Tensor(Of T)

类型参数

T

参数

返回

适用于

Truncate<T>(ReadOnlyTensorSpan<T>, TensorSpan<T>)

Source:
TensorExtensions.cs

计算指定张量中数字的元素截断。

public:
generic <typename T>
 where T : System::Numerics::IFloatingPoint<T> static System::Numerics::Tensors::TensorSpan<T> ^ Truncate(System::Numerics::Tensors::ReadOnlyTensorSpan<T> % x, System::Numerics::Tensors::TensorSpan<T> % destination);
public static ref readonly System.Numerics.Tensors.TensorSpan<T> Truncate<T> (scoped in System.Numerics.Tensors.ReadOnlyTensorSpan<T> x, in System.Numerics.Tensors.TensorSpan<T> destination) where T : System.Numerics.IFloatingPoint<T>;
static member Truncate : ReadOnlyTensorSpan * TensorSpan -> TensorSpan (requires 'T :> System.Numerics.IFloatingPoint<'T>)
Public Function Truncate(Of T As IFloatingPoint(Of T)) (ByRef x As ReadOnlyTensorSpan(Of T), ByRef destination As TensorSpan(Of T)) As TensorSpan(Of T)

类型参数

T

参数

destination
TensorSpan<T>

返回

适用于