| |
| |
| ``mx.symbol.broadcast_sub`` |
| ====================================================== |
| |
| Description |
| ---------------------- |
| |
| Returns element-wise difference of the input arrays with broadcasting. |
| |
| `broadcast_minus` is an alias to the function `broadcast_sub`. |
| |
| |
| **Example**:: |
| |
| |
| x = [[ 1., 1., 1.], |
| [ 1., 1., 1.]] |
| |
| y = [[ 0.], |
| [ 1.]] |
| |
| broadcast_sub(x, y) = [[ 1., 1., 1.], |
| [ 0., 0., 0.]] |
| |
| broadcast_minus(x, y) = [[ 1., 1., 1.], |
| [ 0., 0., 0.]] |
| |
| Supported sparse operations: |
| |
| broadcast_sub/minus(csr, dense(1D)) = dense |
| broadcast_sub/minus(dense(1D), csr) = dense |
| |
| |
| |
| |
| Usage |
| ---------- |
| |
| .. code:: r |
| |
| mx.symbol.broadcast_sub(...) |
| |
| Arguments |
| ------------------ |
| |
| +----------------------------------------+------------------------------------------------------------+ |
| | Argument | Description | |
| +========================================+============================================================+ |
| | ``lhs`` | NDArray-or-Symbol. | |
| | | | |
| | | First input to the function | |
| +----------------------------------------+------------------------------------------------------------+ |
| | ``rhs`` | NDArray-or-Symbol. | |
| | | | |
| | | Second input to the function | |
| +----------------------------------------+------------------------------------------------------------+ |
| | ``name`` | string, optional. | |
| | | | |
| | | Name of the resulting symbol. | |
| +----------------------------------------+------------------------------------------------------------+ |
| |
| Value |
| ---------- |
| |
| ``out`` The result mx.symbol |
| |
| |
| Link to Source Code: http://github.com/apache/incubator-mxnet/blob/1.6.0/src/operator/tensor/elemwise_binary_broadcast_op_basic.cc#L106 |
| |