com.intel.analytics.bigdl.tensor.SparseTensorMath.scala Maven / Gradle / Ivy
/*
* Copyright 2016 The BigDL Authors.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package com.intel.analytics.bigdl.tensor
import com.intel.analytics.bigdl.tensor.TensorNumericMath._
object SparseTensorMath {
def vdot[@specialized(Float, Double) T](
vec1: DenseTensor[T],
vec2: SparseTensor[T]): T = {
SparseTensorBLAS.vdot(vec1, vec2)
}
def addmv[@specialized(Float, Double) T](
r : Tensor[T],
beta : T,
t : Tensor[T],
alpha : T,
mat : Tensor[T],
vec : Tensor[T])(implicit ev: TensorNumeric[T]): Tensor[T] = {
require(mat.nDimension() == 2 && vec.nDimension() == 1)
require(mat.size(2) == vec.size(1))
require(t.nDimension() == 1)
require(t.size(1) == mat.size(1))
if(!r.eq(t)) {
r.resizeAs(t).copy(t)
}
SparseTensorBLAS.coomv(alpha, mat, vec, beta, r)
r
}
// res = beta * mat3 + alpha * mat1 * mat2
def addmm[@specialized(Float, Double) T](
res: Tensor[T],
beta: T,
mat3: Tensor[T],
alpha: T,
mat1: Tensor[T],
mat2: Tensor[T]
)(implicit ev: TensorNumeric[T]) : Tensor[T] = {
require(mat1.dim() == 2 && mat2.dim() == 2 && mat3.dim() == 2)
require(mat3.size(1) == mat1.size(1) && mat3.size(2) == mat2.size(2))
if(!res.eq(mat3)) {
res.resizeAs(mat3).copy(mat3)
}
SparseTensorBLAS.coomm(alpha, mat1, mat2, beta, res)
res
}
}