/// | \f$E[d_1,\dots,d_n]\f$ | The tensor \f$T'\f$, where \f$T'[i_1,\dots,i_{m-1},i_m,i_{m+1},\dots,i_n] = 1\f$ if \f$T[i_1,\dots,i_{m-1},i_{m+1},\dots,i_n] = i_m\f$, else \f$0\f$. However, \f$T'\f$ is undefined if any non-integral value or any out-of-bounds value is detected in the input tensor. |
/// (Note that `below` and `above` here refer respectively to lower- or higher-numbered coordinate indices, and numbering starts at the upper-left corner;
/// thus inserting a row "below" actually inserts it at the "top" of the matrix.)
///
classPad:publicRequiresTensorViewArgs
classPad:publicutil::RequiresTensorViewArgs
{
public:
/// \brief Constructs a generic padding operation.
/// | \f$E[\textit{delete}(A,d_1,\dots,d_n)]\f$ | The tensor \f$T\f$, where \f$T\f$ is the input tensor with the `reduction_axes` \f$A\f$ eliminated by reduction. |
/// | \f$E[d'_1,\dots,d'_n]\f$ | The tensor \f$T\f$, where \f$T[i_1,\dots,i_n] = \mathit{reduce}(\mathit{reduction\_function},\mathit{arg\_init},V)\f$ where \f$V\f$ is the set of values in the input tensor within the window defined by the lower bound \f$(s_1i_1,\dots,s_ni_n)\f$ and the noninclusive upper bound \f$(s_1i_1 + w_1,\dots,s_ni_n + w_n)\f$. |
/// | \f$E[d_1,\dots,d_n]\f$ | The tensor \f$T\f$ where \f$T[i_1,\dots,i_n] = \texttt{arg1}[j_1,\dots,j_n]\f$ if \f$j_1,\dots,j_n\f$ is in bounds for `arg1` and for all \f$m\f$, \f$i_m = l_m + j_m s_m\f$, otherwise \f$\texttt{arg0}[i_1,\dots,i_n]\f$. |
/// | \f$E[d_1,\dots,d_n]\f$ | The tensor \f$T\f$, where \f$T[i_1,\dots,i_n] = \texttt{arg}[j_1,\dots,j_n]\f$ and \f$j_k = d_k - i_k - 1\f$ if axis \f$k\f$ is in the reverse set; else \f$j_k = i_k\f$. |
/// | \f$N[\textit{delete}(A,d_1,\dots,d_n)]\f$ | The tensor \f$T\f$, where \f$T\f$ is the input tensor with the `reduction_axes` \f$A\f$ eliminated by summation. |
/// \brief Abstract base class for elementwise binary operations, i.e., operations where the same
/// scalar binary operation is applied to each corresponding pair of elements in two same-shaped
/// input tensors.
///
/// For example, if the underlying operation (determined by the subclass) is \f$\mathit{op}(x,y)\f$, the input tensors
/// \f$[[x_0,y_0],[z_0,w_0]]\f$ and \f$[[x_1,y_1],[z_1,w_1]]\f$ will be mapped to \f$[[\mathit{op}(x_0,x_1),\mathit{op}(y_0,y_1)],[\mathit{op}(z_0,z_1),\mathit{op}(w_0,w_1)]]\f$.
/// | `arg0` | \f$E_0[d_1,\dots,d_n]~(n \geq 0)\f$ | A tensor of any shape. Subclasses may impose restrictions on the element type \f$E_0\f$. |
/// | `arg1` | \f$E_1[d_1,\dots,d_n]~(n \geq 0)\f$ | A tensor of the same shape as `arg0`. Subclasses may impose restrictions on the element type \f$E_1\f$. |
/// | \f$E_2[d_1,\dots,d_n]\f$ | The tensor \f$T\f$, where \f$T[i_1,\dots,i_n] = \mathit{op}(\texttt{arg0}[i_1,\dots,i_n],\texttt{arg1}[i_1,\dots,i_n])\f$. This will always have the same shape as the input tensors, but subclasses must determine the element type \f$E_2\f$. |
/// \brief Abstract base class for elementwise binary arithmetic operations, i.e., operations where the same
/// scalar binary arithmetic operation is applied to each corresponding pair of elements in two same-shaped
/// input tensors.
///
/// For example, if the underlying arithmetic operation (determined by the subclass) is \f$\mathit{op}(x,y)\f$, the input tensors
/// \f$[[x_0,y_0],[z_0,w_0]]\f$ and \f$[[x_1,y_1],[z_1,w_1]]\f$ will be mapped to \f$[[\mathit{op}(x_0,x_1),\mathit{op}(y_0,y_1)],[\mathit{op}(z_0,z_1),\mathit{op}(w_0,w_1)]]\f$.
/// | \f$N[d_1,\dots,d_n]\f$ | The tensor \f$T\f$, where \f$T[i_1,\dots,i_n] = \mathit{op}(\texttt{arg0}[i_1,\dots,i_n],\texttt{arg1}[i_1,\dots,i_n])\f$. This will always have the same shape and element type as the input tensors. |
/// \brief Abstract base class for elementwise binary comparison operations, i.e., operations where the same
/// scalar binary comparison operation is applied to each corresponding pair of elements in two same-shaped
/// input tensors.
///
/// For example, if the underlying comparison operation (determined by the subclass) is \f$\mathit{op}(x,y)\f$, the input tensors
/// \f$[[x_0,y_0],[z_0,w_0]]\f$ and \f$[[x_1,y_1],[z_1,w_1]]\f$ will be mapped to \f$[[\mathit{op}(x_0,x_1),\mathit{op}(y_0,y_1)],[\mathit{op}(z_0,z_1),\mathit{op}(w_0,w_1)]]\f$.
/// | \f$\texttt{bool}[d_1,\dots,d_n]\f$ | The tensor \f$T\f$, where \f$T[i_1,\dots,i_n] = \mathit{op}(\texttt{arg0}[i_1,\dots,i_n],\texttt{arg1}[i_1,\dots,i_n])\f$. This will always have the same shape as the input tensors, and the element type `bool`. |
/// | \f$E'[d_1,\dots,d_n]\f$ | The tensor \f$T\f$, where \f$T[i_1,\dots,i_n] = \mathit{op}(\texttt{arg}[i_1,\dots,i_n])\f$. This will always have the same shape as the input tensor, but subclasses must determine the element type \f$E'\f$. |
/// | \f$N[d_1,\dots,d_n]\f$ | The tensor \f$T\f$, where \f$T[i_1,\dots,i_n] = \mathit{op}(\texttt{arg}[i_1,\dots,i_n])\f$. This will always have the same shape and element type as the input tensor. |
