2019-06-06 14:21:15 +02:00
|
|
|
/*******************************************************************************
|
|
|
|
* Copyright (c) 2015-2018 Skymind, Inc.
|
|
|
|
*
|
|
|
|
* This program and the accompanying materials are made available under the
|
|
|
|
* terms of the Apache License, Version 2.0 which is available at
|
|
|
|
* https://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.
|
|
|
|
*
|
|
|
|
* SPDX-License-Identifier: Apache-2.0
|
|
|
|
******************************************************************************/
|
|
|
|
|
|
|
|
//
|
|
|
|
// @author raver119@gmail.com
|
|
|
|
//
|
|
|
|
|
|
|
|
#include <ops/declarable/helpers/top_k.h>
|
|
|
|
#include <MmulHelper.h>
|
|
|
|
#include <NDArrayFactory.h>
|
|
|
|
#include <Status.h>
|
2019-12-20 15:56:28 +01:00
|
|
|
#include <execution/Threads.h>
|
|
|
|
#include <execution/Threads.h>
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
namespace nd4j {
|
|
|
|
namespace ops {
|
|
|
|
namespace helpers {
|
|
|
|
|
|
|
|
template <typename T>
|
|
|
|
static void swapRows_(NDArray* matrix, int theFirst, int theSecond) {
|
|
|
|
|
|
|
|
if (theFirst != theSecond)
|
|
|
|
for (int i = 0; i < matrix->columns(); i++) {
|
2019-12-20 15:56:28 +01:00
|
|
|
math::nd4j_swap(matrix->t<T>(theFirst, i), matrix->t<T>(theSecond, i));
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
BUILD_SINGLE_TEMPLATE(template void swapRows_, (NDArray* matrix, int theFirst, int theSecond), FLOAT_TYPES);
|
|
|
|
|
2019-12-20 15:56:28 +01:00
|
|
|
template <typename T>
|
|
|
|
static void swapRows(T* matrixBuf, Nd4jLong* matrixShape, Nd4jLong theFirst, Nd4jLong theSecond) {
|
|
|
|
if (theFirst != theSecond) {
|
|
|
|
auto n = shape::sizeAt(matrixShape, -1);
|
|
|
|
|
|
|
|
auto loop = PRAGMA_THREADS_FOR {
|
2020-02-20 09:43:26 +01:00
|
|
|
for (auto i = start; i < stop; i++) {
|
2019-12-20 15:56:28 +01:00
|
|
|
Nd4jLong theFirstPos[] = {theFirst, i};
|
|
|
|
Nd4jLong theSecondPos[] = {theSecond, i};
|
|
|
|
auto theFirstIndex = shape::getOffset(matrixShape, theFirstPos, 0);
|
|
|
|
auto theSecondIndex = shape::getOffset(matrixShape, theSecondPos, 0);
|
|
|
|
math::nd4j_swap(matrixBuf[theFirstIndex], matrixBuf[theSecondIndex]);
|
|
|
|
}
|
|
|
|
};
|
|
|
|
|
|
|
|
samediff::Threads::parallel_tad(loop, 0, n, 1);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2019-06-06 14:21:15 +02:00
|
|
|
void swapRows(NDArray* matrix, int theFirst, int theSecond) {
|
|
|
|
BUILD_SINGLE_SELECTOR(matrix->dataType(), swapRows_, (matrix, theFirst, theSecond), FLOAT_TYPES);
|
|
|
|
}
|
|
|
|
|
|
|
|
template <typename T>
|
|
|
|
static void invertLowerMatrix_(NDArray* inputMatrix, NDArray* invertedMatrix) {
|
|
|
|
int n = inputMatrix->rows();
|
2020-02-28 09:37:26 +01:00
|
|
|
invertedMatrix->setIdentity();
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
if (inputMatrix->isIdentityMatrix()) return;
|
|
|
|
|
2020-02-28 09:37:26 +01:00
|
|
|
auto invertDiagonals = PRAGMA_THREADS_FOR {
|
|
|
|
for (int i = start; i < stop; i += increment)
|
|
|
|
invertedMatrix->t<T>(i, i) /= inputMatrix->t<T>(i, i);
|
|
|
|
};
|
2019-06-06 14:21:15 +02:00
|
|
|
|
2020-02-28 09:37:26 +01:00
|
|
|
auto invertSubDiagonals = PRAGMA_THREADS_FOR {
|
|
|
|
for (int i = start; i < stop; i += increment)
|
|
|
|
invertedMatrix->t<T>(i, i - 1) -= (inputMatrix->t<T>(i, i - 1) * invertedMatrix->t<T>(i - 1, i - 1) / inputMatrix->t<T>(i, i));
|
|
|
|
};
|
|
|
|
|
|
|
|
samediff::Threads::parallel_for(invertDiagonals, 0, n, 1);
|
|
|
|
samediff::Threads::parallel_for(invertSubDiagonals, 1, n, 1);
|
|
|
|
|
|
|
|
// PRAGMA_OMP_PARALLEL_FOR_SIMD
|
|
|
|
for (int i = 1; i < n; i++) {
|
|
|
|
for (int j = 0; j < i - 1 ; j++)
|
2019-06-06 14:21:15 +02:00
|
|
|
for (int k = 0; k < i; k++)
|
2020-02-28 09:37:26 +01:00
|
|
|
invertedMatrix->t<T>(i, j) -= ((invertedMatrix->t<T>(k, j) * inputMatrix->t<T>(i, k) / inputMatrix->t<T>(i, i)));
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
2020-02-28 09:37:26 +01:00
|
|
|
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
BUILD_SINGLE_TEMPLATE(template void invertLowerMatrix_, (NDArray* inputMatrix, NDArray* invertedMatrix);, FLOAT_TYPES);
|
|
|
|
|
|
|
|
void invertLowerMatrix(NDArray* inputMatrix, NDArray* invertedMatrix) {
|
|
|
|
BUILD_SINGLE_SELECTOR(inputMatrix->dataType(), invertLowerMatrix_, (inputMatrix, invertedMatrix), FLOAT_TYPES);
|
|
|
|
}
|
|
|
|
|
|
|
|
template <typename T>
|
|
|
|
static void _invertUpperMatrix(NDArray* inputMatrix, NDArray* invertedMatrix) {
|
|
|
|
int n = inputMatrix->rows();
|
|
|
|
invertedMatrix->setIdentity();
|
|
|
|
|
|
|
|
if (inputMatrix->isIdentityMatrix()) { // the inverse for I is I
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2020-02-28 09:37:26 +01:00
|
|
|
auto invertDiagonals = PRAGMA_THREADS_FOR {
|
|
|
|
for (auto i = start; i < stop; i += increment)
|
|
|
|
invertedMatrix->t<T>(i, i) /= inputMatrix->t<T>(i, i);
|
|
|
|
};
|
2019-06-06 14:21:15 +02:00
|
|
|
|
2019-09-09 15:27:45 +02:00
|
|
|
//PRAGMA_OMP_PARALLEL_FOR_IF(n > Environment::getInstance()->elementwiseThreshold())
|
2020-02-28 09:37:26 +01:00
|
|
|
auto invertUpDiagonals = PRAGMA_THREADS_FOR {
|
|
|
|
for (auto i = start; i < stop; i += increment)
|
|
|
|
invertedMatrix->t<T>(i, i + 1) -= (inputMatrix->t<T>(i, i + 1) * invertedMatrix->t<T>(i + 1, i + 1) /
|
|
|
|
inputMatrix->t<T>(i, i));
|
|
|
|
};
|
|
|
|
|
|
|
|
samediff::Threads::parallel_for(invertDiagonals, 0, n, 1);
|
|
|
|
samediff::Threads::parallel_for(invertUpDiagonals, 0, n - 1, 1);
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
// PRAGMA_OMP_PARALLEL_FOR_SIMD
|
2020-02-28 09:37:26 +01:00
|
|
|
for (auto i = n - 2; i >= 0; i--) {
|
|
|
|
for (auto j = i + 2; j < n; j++)
|
|
|
|
for (auto k = i; k < n; k++)
|
2019-06-06 14:21:15 +02:00
|
|
|
invertedMatrix->t<T>(i, j) -= ((invertedMatrix->t<T>(k, j) * inputMatrix->t<T>(i, k) / inputMatrix->t<T>(i, i)));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
BUILD_SINGLE_TEMPLATE(template void _invertUpperMatrix, (NDArray* inputMatrix, NDArray* invertedMatrix);, FLOAT_TYPES);
|
|
|
|
|
|
|
|
void invertUpperMatrix(NDArray* inputMatrix, NDArray* invertedMatrix) {
|
|
|
|
BUILD_SINGLE_SELECTOR(inputMatrix->dataType(), _invertUpperMatrix, (inputMatrix, invertedMatrix), FLOAT_TYPES);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2019-12-20 15:56:28 +01:00
|
|
|
template <typename T, typename I>
|
2019-08-24 15:59:30 +02:00
|
|
|
static NDArray lup_(LaunchContext *context, NDArray* input, NDArray* compound, NDArray* permutation) {
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
const int rowNum = input->rows();
|
|
|
|
const int columnNum = input->columns();
|
|
|
|
|
|
|
|
NDArray determinant = NDArrayFactory::create<T>(1.f);
|
|
|
|
NDArray compoundMatrix = *input; // copy
|
2019-08-24 15:59:30 +02:00
|
|
|
NDArray permutationMatrix(input, false, context); // has same shape as input and contiguous strides
|
2019-06-06 14:21:15 +02:00
|
|
|
permutationMatrix.setIdentity();
|
|
|
|
|
|
|
|
T pivotValue; // = T(0.0);
|
|
|
|
int pivot; // = -1;
|
|
|
|
int swapCount = 0;
|
|
|
|
|
|
|
|
for(int i = 0; i < rowNum; i++ ) {
|
|
|
|
pivotValue = T(0.0);
|
|
|
|
pivot = -1;
|
2019-09-09 15:27:45 +02:00
|
|
|
//PRAGMA_OMP_PARALLEL_FOR //_ARGS(firstprivate(pivot,pivotValue))
|
2019-06-06 14:21:15 +02:00
|
|
|
for(int rowCounter = i; rowCounter < rowNum; rowCounter++ ) {
|
2019-07-12 10:51:51 +02:00
|
|
|
if (nd4j::math::nd4j_abs(compoundMatrix.t<T>(rowCounter, i)) > pivotValue) {
|
|
|
|
pivotValue = nd4j::math::nd4j_abs(compoundMatrix.t<T>(rowCounter, i));
|
2019-06-06 14:21:15 +02:00
|
|
|
pivot = rowCounter;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2019-12-20 15:56:28 +01:00
|
|
|
if( pivotValue > DataTypeUtils::min<T>()) {
|
2019-06-06 14:21:15 +02:00
|
|
|
swapRows(&compoundMatrix, pivot, i);
|
|
|
|
swapRows(&permutationMatrix, pivot, i);
|
|
|
|
if (pivot != i)
|
|
|
|
swapCount++;
|
|
|
|
|
|
|
|
for( int j = i + 1; j < rowNum; j++ ) {
|
2019-07-12 10:51:51 +02:00
|
|
|
compoundMatrix.t<T>(j, i) /= compoundMatrix.t<T>(i, i);
|
2019-09-09 15:27:45 +02:00
|
|
|
//PRAGMA_OMP_PARALLEL_FOR
|
2019-06-06 14:21:15 +02:00
|
|
|
for( int k = i + 1; k < rowNum; k++ ) {
|
2019-07-12 10:51:51 +02:00
|
|
|
compoundMatrix.t<T>(j, k) -= compoundMatrix.t<T>(j, i) * compoundMatrix.t<T>(i, k);
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
for (int e = 0; e < rowNum; e++) {
|
|
|
|
// nd4j_printf("Compound matrix diag %i %f.\n", e, (*compoundMatrix)(e, e));
|
|
|
|
determinant *= compoundMatrix.e<T>(e, e);
|
|
|
|
}
|
|
|
|
if (swapCount % 2) determinant = -determinant;
|
|
|
|
if (compound != nullptr)
|
|
|
|
compound->assign(compoundMatrix);
|
2019-12-20 15:56:28 +01:00
|
|
|
if (permutation != nullptr) {
|
|
|
|
auto permutaionVector = NDArrayFactory::create('c', {rowNum}, DataTypeUtils::fromT<I>(), input->getContext());
|
|
|
|
for (auto i = 0; i < rowNum; i++) {
|
|
|
|
for (auto j = 0; j < columnNum; j++) {
|
|
|
|
if (permutationMatrix.t<T>(i, j) != 0) {
|
|
|
|
permutaionVector.template t<I>(i) = j;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (permutationMatrix.isSameShape(permutation))
|
|
|
|
permutation->assign(permutationMatrix);
|
|
|
|
else if (permutation->isSameShape(permutaionVector)) {
|
|
|
|
permutation->assign(permutaionVector);
|
|
|
|
}
|
|
|
|
}
|
2019-06-06 14:21:15 +02:00
|
|
|
return determinant;
|
|
|
|
}
|
|
|
|
|
2019-12-20 15:56:28 +01:00
|
|
|
BUILD_DOUBLE_TEMPLATE(template NDArray lup_, (LaunchContext *context, NDArray* input, NDArray* output, NDArray* permutation), FLOAT_TYPES, INDEXING_TYPES);
|
|
|
|
/*
|
|
|
|
* lu decomposition with naive algorithm with partial pivoting
|
|
|
|
* */
|
|
|
|
template <typename T, typename I>
|
|
|
|
static I argmaxCol(I column, T* compoundBuffer, Nd4jLong* compoundShape) {
|
|
|
|
auto rowNum = shape::sizeAt(compoundShape, 0);
|
|
|
|
Nd4jLong xInitial[] = {column, column};
|
|
|
|
auto xInitialIndex = shape::getOffset(compoundShape, xInitial, 0);
|
|
|
|
auto maxValue = T(0); //nd4j::math::nd4j_abs(compoundBuffer[xInitialIndex]);
|
|
|
|
auto result = -1;
|
|
|
|
//auto loop = PRAGMA_THREADS_FOR {
|
|
|
|
auto start = column, stop = rowNum, increment = 1;
|
2020-02-20 09:43:26 +01:00
|
|
|
for (auto rowCounter = start; rowCounter < stop; rowCounter++) {
|
2019-12-20 15:56:28 +01:00
|
|
|
Nd4jLong xPos[] = {rowCounter, column};
|
|
|
|
auto xIndex = shape::getOffset(compoundShape, xPos, 0);
|
|
|
|
if (nd4j::math::nd4j_abs(compoundBuffer[xIndex]) > maxValue) {
|
|
|
|
maxValue = nd4j::math::nd4j_max(maxValue, nd4j::math::nd4j_abs(compoundBuffer[xIndex]));
|
|
|
|
result = rowCounter;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
//};
|
|
|
|
//samediff::Threads::parallel_for(loop, column, rowNum, 1);
|
|
|
|
return result;
|
|
|
|
}
|
2019-06-06 14:21:15 +02:00
|
|
|
|
2019-12-20 15:56:28 +01:00
|
|
|
template <typename T>
|
|
|
|
void processColumns(int currentRow, int rowNum, T* compoundBuf, Nd4jLong* compoundShape) {
|
|
|
|
Nd4jLong xDiag[] = {currentRow, currentRow};
|
|
|
|
auto diagIndex = shape::getOffset(compoundShape, xDiag, 0);
|
|
|
|
auto loop = PRAGMA_THREADS_FOR {
|
2020-02-20 09:43:26 +01:00
|
|
|
for (auto j = start; j < stop; j++) {
|
2019-12-20 15:56:28 +01:00
|
|
|
Nd4jLong xRow[] = {j, currentRow};
|
|
|
|
auto rowIndex = shape::getOffset(compoundShape, xRow, 0);
|
|
|
|
compoundBuf[rowIndex] /= compoundBuf[diagIndex]; //output->t<T>(i, i);
|
|
|
|
for (int k = currentRow + 1; k < rowNum; k++) {
|
|
|
|
Nd4jLong yRow[] = {j, k};
|
|
|
|
Nd4jLong yCol[] = {currentRow, k};
|
|
|
|
auto rowIndexY = shape::getOffset(compoundShape, yRow, 0);
|
|
|
|
auto colIndex = shape::getOffset(compoundShape, yCol, 0);
|
|
|
|
compoundBuf[rowIndexY] -= compoundBuf[rowIndex] * compoundBuf[colIndex];
|
|
|
|
}
|
|
|
|
}
|
|
|
|
};
|
|
|
|
samediff::Threads::parallel_tad(loop, currentRow + 1, rowNum, 1);
|
|
|
|
}
|
2019-06-06 14:21:15 +02:00
|
|
|
|
2020-02-04 06:59:11 +01:00
|
|
|
template <typename T>
|
|
|
|
static void doolitleLU(LaunchContext* context, NDArray* compound, Nd4jLong rowNum) {
|
|
|
|
auto input = compound->dup();
|
|
|
|
compound->nullify();
|
|
|
|
|
|
|
|
// Decomposing matrix into Upper and Lower
|
|
|
|
// triangular matrix
|
|
|
|
for (auto i = 0; i < rowNum; i++) {
|
|
|
|
|
|
|
|
// Upper Triangular
|
|
|
|
for (auto k = i; k < rowNum; k++) {
|
|
|
|
|
|
|
|
// Summation of L(i, j) * U(j, k)
|
|
|
|
int sum = 0;
|
|
|
|
for (int j = 0; j < i; j++)
|
|
|
|
sum += compound->t<T>(i,j) * compound->t<T>(j,k);
|
|
|
|
|
|
|
|
// Evaluating U(i, k)
|
|
|
|
compound->t<T>(i, k) = input.t<T>(i, k) - sum;
|
|
|
|
}
|
|
|
|
|
|
|
|
// Lower Triangular
|
|
|
|
for (int k = i + 1; k < rowNum; k++) {
|
|
|
|
// Summation of L(k, j) * U(j, i)
|
|
|
|
int sum = 0;
|
|
|
|
for (int j = 0; j < i; j++)
|
|
|
|
sum += compound->t<T>(k,j) * compound->t<T>(j, i);
|
|
|
|
|
|
|
|
// Evaluating L(k, i)
|
|
|
|
compound->t<T>(k, i) = (input.t<T>(k, i) - sum) / compound->t<T>(i,i);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2019-12-20 15:56:28 +01:00
|
|
|
template <typename T, typename I>
|
|
|
|
static void luNN_(LaunchContext *context, NDArray* compound, NDArray* permutation, Nd4jLong rowNum) {
|
|
|
|
|
|
|
|
//const int rowNum = compound->rows();
|
|
|
|
// const int columnNum = output->columns();
|
2020-02-04 06:59:11 +01:00
|
|
|
if (permutation) { // LUP algorithm
|
|
|
|
permutation->linspace(0);
|
|
|
|
auto permutationBuf = permutation->bufferAsT<I>(); //dataBuffer()->primaryAsT<I>();
|
|
|
|
auto compoundBuf = compound->bufferAsT<T>();
|
|
|
|
auto compoundShape = compound->shapeInfo();
|
|
|
|
auto permutationShape = permutation->shapeInfo();
|
|
|
|
for (auto i = 0; i < rowNum - 1; i++) {
|
|
|
|
auto pivotIndex = argmaxCol(i, compoundBuf, compoundShape);
|
|
|
|
if (pivotIndex < 0) {
|
|
|
|
throw std::runtime_error("helpers::luNN_: input matrix is singular.");
|
|
|
|
}
|
|
|
|
math::nd4j_swap(permutationBuf[shape::getIndexOffset(i, permutationShape)],
|
|
|
|
permutationBuf[shape::getIndexOffset(pivotIndex, permutationShape)]);
|
|
|
|
swapRows(compoundBuf, compoundShape, i, pivotIndex);
|
2019-12-20 15:56:28 +01:00
|
|
|
|
2020-02-04 06:59:11 +01:00
|
|
|
processColumns(i, rowNum, compoundBuf, compoundShape);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else { // Doolitle algorithm with LU decomposition
|
|
|
|
doolitleLU<T>(context, compound, rowNum);
|
2019-12-20 15:56:28 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
template <typename T, typename I>
|
|
|
|
static void lu_(LaunchContext * context, NDArray* input, NDArray* output, NDArray* permutationVectors) {
|
|
|
|
auto n = input->sizeAt(-1);
|
|
|
|
|
|
|
|
output->assign(input); // fill up output tensor with zeros
|
2019-12-20 20:35:39 +01:00
|
|
|
ResultSet outputs = output->allTensorsAlongDimension({-2, -1});
|
2020-02-04 06:59:11 +01:00
|
|
|
ResultSet permutations;
|
|
|
|
if (permutationVectors)
|
|
|
|
permutations = permutationVectors->allTensorsAlongDimension({-1});
|
|
|
|
|
2019-12-20 15:56:28 +01:00
|
|
|
auto loop = PRAGMA_THREADS_FOR {
|
2020-02-20 09:43:26 +01:00
|
|
|
for (auto i = start; i < stop; i++) {
|
2020-02-04 06:59:11 +01:00
|
|
|
luNN_<T, I>(context, outputs.at(i), permutationVectors?permutations.at(i):nullptr, n);
|
2019-12-20 15:56:28 +01:00
|
|
|
}
|
|
|
|
};
|
2019-12-20 20:35:39 +01:00
|
|
|
samediff::Threads::parallel_for(loop, 0, outputs.size(), 1);
|
2019-12-20 15:56:28 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
void lu(LaunchContext *context, NDArray* input, NDArray* output, NDArray* permutation) {
|
2020-02-04 06:59:11 +01:00
|
|
|
BUILD_DOUBLE_SELECTOR(input->dataType(), permutation?permutation->dataType():DataType::INT32, lu_, (context, input, output, permutation), FLOAT_TYPES, INDEXING_TYPES);
|
2019-12-20 15:56:28 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
// BUILD_DOUBLE_TEMPLATE(template NDArray lu_, (LaunchContext *context, NDArray* input, NDArray* output, NDArray* permutation), FLOAT_TYPES, INDEXING_TYPES);
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
template <typename T>
|
2019-08-24 15:59:30 +02:00
|
|
|
static int determinant_(LaunchContext *context, NDArray* input, NDArray* output) {
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
Nd4jLong n = input->sizeAt(-1);
|
|
|
|
Nd4jLong n2 = n * n;
|
|
|
|
|
2019-08-24 15:59:30 +02:00
|
|
|
auto matrix = NDArrayFactory::create(input->ordering(), {n, n}, input->dataType(), context); //, block.getWorkspace());
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
for (int e = 0; e < output->lengthOf(); e++) {
|
|
|
|
for (int k = e * n2, row = 0; k < (e + 1) * n2; ++k, ++row)
|
|
|
|
matrix.p(row, input->e<T>(k));
|
2019-12-20 15:56:28 +01:00
|
|
|
output->p(e, lup_<T, int>(context, &matrix, (NDArray*)nullptr, (NDArray*)nullptr));
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
return Status::OK();
|
|
|
|
}
|
|
|
|
|
|
|
|
int determinant(nd4j::LaunchContext * context, NDArray* input, NDArray* output) {
|
2019-08-24 15:59:30 +02:00
|
|
|
BUILD_SINGLE_SELECTOR(input->dataType(), return determinant_, (context, input, output), FLOAT_TYPES);
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
template <typename T>
|
2019-08-24 15:59:30 +02:00
|
|
|
int logAbsDeterminant_(LaunchContext *context, NDArray* input, NDArray* output) {
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
Nd4jLong n = input->sizeAt(-1);
|
|
|
|
Nd4jLong n2 = n * n;
|
|
|
|
|
2019-08-24 15:59:30 +02:00
|
|
|
NDArray matrix = NDArrayFactory::create(input->ordering(), {n, n}, input->dataType(), context); //, block.getWorkspace());
|
2019-06-06 14:21:15 +02:00
|
|
|
for (int e = 0; e < output->lengthOf(); e++) {
|
|
|
|
for (int k = e * n2, row = 0; k < (e + 1) * n2; ++k, ++row) {
|
|
|
|
matrix.p(row, input->e<T>(k));
|
|
|
|
}
|
2019-12-20 15:56:28 +01:00
|
|
|
NDArray det = lup_<T, int>(context, &matrix, (NDArray*)nullptr, (NDArray*)nullptr);
|
2019-06-06 14:21:15 +02:00
|
|
|
if (det.e<T>(0) != 0.f)
|
|
|
|
output->p(e, nd4j::math::nd4j_log<T,T>(nd4j::math::nd4j_abs(det.t<T>(0))));
|
|
|
|
}
|
|
|
|
|
|
|
|
return ND4J_STATUS_OK;
|
|
|
|
}
|
|
|
|
|
2019-07-12 10:51:51 +02:00
|
|
|
int logAbsDeterminant(nd4j::LaunchContext * context, NDArray* input, NDArray* output) {
|
2019-08-24 15:59:30 +02:00
|
|
|
BUILD_SINGLE_SELECTOR(input->dataType(), return logAbsDeterminant_, (context, input, output), FLOAT_TYPES);
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
template <typename T>
|
2019-08-24 15:59:30 +02:00
|
|
|
static int inverse_(LaunchContext *context, NDArray* input, NDArray* output) {
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
auto n = input->sizeAt(-1);
|
|
|
|
auto n2 = n * n;
|
|
|
|
auto totalCount = output->lengthOf() / n2;
|
|
|
|
|
|
|
|
output->assign(0.f); // fill up output tensor with zeros
|
2019-08-24 15:59:30 +02:00
|
|
|
auto matrix = NDArrayFactory::create('c', {n, n}, DataTypeUtils::fromT<T>(), context); //, block.getWorkspace());
|
|
|
|
auto compound = NDArrayFactory::create('c', {n, n}, DataTypeUtils::fromT<T>(), context); //, block.getWorkspace());
|
|
|
|
auto permutation = NDArrayFactory::create('c', {n, n}, DataTypeUtils::fromT<T>(), context);
|
|
|
|
auto lowerMatrix = NDArrayFactory::create('c', {n, n}, DataTypeUtils::fromT<T>(), context);
|
|
|
|
auto upperMatrix = NDArrayFactory::create('c', {n, n}, DataTypeUtils::fromT<T>(), context);
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
for (int e = 0; e < totalCount; e++) {
|
|
|
|
if (e)
|
|
|
|
matrix.assign(0.f);
|
|
|
|
|
|
|
|
for (int k = e * n2, row = 0; k < (e + 1) * n2; k++) {
|
|
|
|
matrix.p(row++, input->e<T>(k));
|
|
|
|
}
|
2019-12-20 15:56:28 +01:00
|
|
|
T det = lup_<T, int>(context, &matrix, &compound, &permutation).template e<T>(0);
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
// FIXME: and how this is going to work on float16?
|
2019-07-12 10:51:51 +02:00
|
|
|
if (nd4j::math::nd4j_abs<T>(det) < T(0.000001)) {
|
2019-06-06 14:21:15 +02:00
|
|
|
nd4j_printf("matrix_inverse: The matrix %i has no inverse due determinant is %lf. Quiting...\n", e, det);
|
|
|
|
matrix.printIndexedBuffer("Wrong matrix");
|
|
|
|
return ND4J_STATUS_VALIDATION;
|
|
|
|
}
|
|
|
|
lowerMatrix.setIdentity(); // set up U to identity matrix
|
|
|
|
for (int k = 1; k < n; k++) { // and then put all values under main diagonal on to it
|
|
|
|
for (int j = 0; j < k; j++)
|
2019-07-12 10:51:51 +02:00
|
|
|
lowerMatrix.template t<T>(k, j) = compound.template t<T>(k, j);
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
upperMatrix.setIdentity(); // set up U to identity matrix
|
|
|
|
for (int k = 0; k < n; k++) { // and then put all values under main diagonal on to it
|
|
|
|
for (int j = k; j < n; j++)
|
2019-07-12 10:51:51 +02:00
|
|
|
upperMatrix.template t<T>(k, j) = compound.template e<T>(k, j);
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
invertUpperMatrix(&upperMatrix, &matrix);
|
|
|
|
|
|
|
|
invertLowerMatrix(&lowerMatrix, &upperMatrix);
|
|
|
|
|
|
|
|
nd4j::MmulHelper::mmul(&matrix, &upperMatrix, &compound, 1.0, 0.0);
|
|
|
|
nd4j::MmulHelper::mmul(&compound, &permutation, &matrix, 1.0, 0.0);
|
|
|
|
for (int k = e * n2, row = 0; k < (e + 1) * n2; k++) {
|
2019-07-12 10:51:51 +02:00
|
|
|
output->t<T>(k) = matrix.template t<T>(row++);
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return Status::OK();
|
|
|
|
}
|
|
|
|
|
2020-02-28 09:37:26 +01:00
|
|
|
template <typename T>
|
|
|
|
static int lowerInverse_(LaunchContext *context, NDArray* input, NDArray* output) {
|
|
|
|
|
|
|
|
auto n = input->sizeAt(-1);
|
|
|
|
auto n2 = n * n;
|
|
|
|
auto totalCount = output->lengthOf() / n2;
|
|
|
|
|
|
|
|
output->assign(0.f); // fill up output tensor with zeros
|
|
|
|
auto matrix = NDArrayFactory::create('c', {n, n}, DataTypeUtils::fromT<T>(), context); //, block.getWorkspace());
|
|
|
|
auto compound = NDArrayFactory::create('c', {n, n}, DataTypeUtils::fromT<T>(), context); //, block.getWorkspace());
|
|
|
|
auto permutation = NDArrayFactory::create('c', {n, n}, DataTypeUtils::fromT<T>(), context);
|
|
|
|
auto lowerMatrix = NDArrayFactory::create('c', {n, n}, DataTypeUtils::fromT<T>(), context);
|
|
|
|
auto upperMatrix = NDArrayFactory::create('c', {n, n}, DataTypeUtils::fromT<T>(), context);
|
|
|
|
|
|
|
|
// auto batchLoop = PRAGMA_THREADS_FOR {
|
|
|
|
for (int e = 0; e < totalCount; e++) {
|
|
|
|
if (e)
|
|
|
|
matrix.assign(0.f);
|
|
|
|
|
|
|
|
for (int k = e * n2, row = 0; k < (e + 1) * n2; k++) {
|
|
|
|
matrix.p(row++, input->e<T>(k));
|
|
|
|
}
|
|
|
|
T det = T(1.f);
|
|
|
|
for (auto i = 0; i < n; i++) {
|
|
|
|
det *= matrix. template t<T>(i, i);
|
|
|
|
}
|
|
|
|
|
|
|
|
// FIXME: and how this is going to work on float16?
|
|
|
|
if (nd4j::math::nd4j_abs<T>(det) < T(0.000001)) {
|
|
|
|
nd4j_printf("matrix_inverse: The matrix %i has no inverse due determinant is %lf. Quiting...\n", e, det);
|
|
|
|
matrix.printIndexedBuffer("Wrong matrix");
|
|
|
|
return ND4J_STATUS_VALIDATION;
|
|
|
|
}
|
|
|
|
lowerMatrix.nullify();
|
|
|
|
invertLowerMatrix(&matrix, &lowerMatrix);
|
|
|
|
|
|
|
|
for (int k = e * n2, row = 0; k < (e + 1) * n2; k++) {
|
|
|
|
output->t<T>(k) = lowerMatrix.template t<T>(row++);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return Status::OK();
|
|
|
|
}
|
|
|
|
|
|
|
|
template <typename T>
|
|
|
|
static int upperInverse_(LaunchContext *context, NDArray* input, NDArray* output) {
|
|
|
|
|
|
|
|
auto n = input->sizeAt(-1);
|
|
|
|
auto n2 = n * n;
|
|
|
|
|
|
|
|
output->nullify(); // fill up output tensor with zeros
|
|
|
|
// auto matrix = NDArrayFactory::create('c', {n, n}, DataTypeUtils::fromT<T>(), context); //, block.getWorkspace());
|
|
|
|
// auto lowerMatrix = NDArrayFactory::create('c', {n, n}, DataTypeUtils::fromT<T>(), context);
|
|
|
|
// auto upperMatrix = NDArrayFactory::create('c', {n, n}, DataTypeUtils::fromT<T>(), context);
|
|
|
|
auto inputPart = input->allTensorsAlongDimension({-2, -1});
|
|
|
|
auto outputPart = output->allTensorsAlongDimension({-2, -1});
|
|
|
|
auto totalCount = outputPart.size(); //lengthOf() / n2;
|
|
|
|
for (int e = 0; e < totalCount; e++) {
|
|
|
|
invertUpperMatrix(inputPart.at(e), outputPart.at(e));
|
|
|
|
}
|
|
|
|
return Status::OK();
|
|
|
|
}
|
|
|
|
|
2019-06-06 14:21:15 +02:00
|
|
|
int inverse(nd4j::LaunchContext * context, NDArray* input, NDArray* output) {
|
2019-08-24 15:59:30 +02:00
|
|
|
BUILD_SINGLE_SELECTOR(input->dataType(), return inverse_, (context, input, output), FLOAT_TYPES);
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
|
2020-02-28 09:37:26 +01:00
|
|
|
int lowerInverseFunctor(nd4j::LaunchContext * context, NDArray* input, NDArray* output) {
|
|
|
|
BUILD_SINGLE_SELECTOR(input->dataType(), return lowerInverse_, (context, input, output), FLOAT_TYPES);
|
|
|
|
}
|
|
|
|
|
|
|
|
int upperInverseFunctor(nd4j::LaunchContext * context, NDArray* input, NDArray* output) {
|
|
|
|
BUILD_SINGLE_SELECTOR(input->dataType(), return upperInverse_, (context, input, output), FLOAT_TYPES);
|
|
|
|
}
|
|
|
|
|
2019-06-06 14:21:15 +02:00
|
|
|
template <typename T>
|
|
|
|
static bool checkCholeskyInput_(nd4j::LaunchContext * context, NDArray const* input) {
|
|
|
|
//std::unique_ptr<NDArray> matrix(NDArrayFactory::create_('c', {n, n}, input->dataType())); //, block.getWorkspace());
|
2019-12-20 20:35:39 +01:00
|
|
|
ResultSet lastMatrixList = input->allTensorsAlongDimension({input->rankOf() - 2, input->rankOf()-1});
|
|
|
|
for (size_t i = 0; i < lastMatrixList.size(); i++) {
|
|
|
|
auto thisMatrix = lastMatrixList.at(i);
|
2019-06-06 14:21:15 +02:00
|
|
|
// check for symmetric
|
|
|
|
for (Nd4jLong r = 0; r < thisMatrix->rows(); r++)
|
|
|
|
for (Nd4jLong c = 0; c < thisMatrix->columns(); c++)
|
2019-12-20 20:35:39 +01:00
|
|
|
if (nd4j::math::nd4j_abs(thisMatrix->e<T>(r, c) - lastMatrixList.at(i)->e<T>(c,r)) > DataTypeUtils::min<T>()) return false;
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
NDArray output = NDArrayFactory::create<T>(0., context);
|
|
|
|
if (ND4J_STATUS_OK != determinant(context, thisMatrix, &output)) return false;
|
|
|
|
if (output.e<T>(0) <= T(0)) return 0;
|
|
|
|
NDArray reversedMatrix(*thisMatrix);
|
|
|
|
if (ND4J_STATUS_OK != inverse(context, thisMatrix, &reversedMatrix)) return false;
|
|
|
|
if (ND4J_STATUS_OK != determinant(context, &reversedMatrix, &output)) return false;
|
|
|
|
if (output.e<T>(0) <= T(0)) return 0;
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
bool checkCholeskyInput(nd4j::LaunchContext * context, NDArray const* input) {
|
|
|
|
BUILD_SINGLE_SELECTOR(input->dataType(), return checkCholeskyInput_, (context, input), FLOAT_TYPES);
|
|
|
|
}
|
|
|
|
|
|
|
|
template <typename T>
|
2019-08-24 15:59:30 +02:00
|
|
|
int cholesky_(LaunchContext *context, NDArray* input, NDArray* output, bool inplace) {
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
auto n = input->sizeAt(-1);
|
|
|
|
auto n2 = n * n;
|
|
|
|
auto totalCount = output->lengthOf() / n2;
|
|
|
|
if (!inplace)
|
|
|
|
output->assign(0.f); // fill up output tensor with zeros only inplace=false
|
|
|
|
|
2019-08-24 15:59:30 +02:00
|
|
|
std::unique_ptr<NDArray> matrix(NDArrayFactory::create_('c', {n, n}, input->dataType(), context)); //, block.getWorkspace());
|
|
|
|
std::unique_ptr<NDArray> lowerMatrix(NDArrayFactory::create_('c',{n, n}, input->dataType(), context));
|
2019-06-06 14:21:15 +02:00
|
|
|
|
|
|
|
for (int e = 0; e < totalCount; e++) {
|
|
|
|
|
|
|
|
// fill up matrix
|
|
|
|
for (int k = e * n2, l = 0; k < (e + 1) * n2; k++) {
|
|
|
|
matrix->p(l++, input->e<T>(k));
|
|
|
|
}
|
|
|
|
//if (e) // from the second loop need to zero matrix
|
|
|
|
lowerMatrix->assign(0.f);
|
|
|
|
|
|
|
|
for (Nd4jLong col = 0; col < n; col++) {
|
|
|
|
for (Nd4jLong row = 0; row < col; row++) {
|
|
|
|
T rowSum = 0;
|
|
|
|
for (Nd4jLong k = 0; k < row; ++k)
|
|
|
|
rowSum += (lowerMatrix->e<T>(col, k) * lowerMatrix->e<T>(row, k));
|
|
|
|
lowerMatrix->p(col, row, (matrix->e<T>(row, col) - rowSum) / lowerMatrix->e<double>(row, row));
|
|
|
|
}
|
|
|
|
T diagonalSum = 0;
|
|
|
|
for (Nd4jLong k = 0; k < col; ++k)
|
|
|
|
diagonalSum += lowerMatrix->e<T>(col, k) * lowerMatrix->e<T>(col, k);
|
|
|
|
lowerMatrix->p(col, col, nd4j::math::nd4j_sqrt<T, T>(matrix->e<T>(col, col) - diagonalSum));
|
|
|
|
//nd4j_printf("%i: ", col);
|
|
|
|
//lowerMatrix->printIndexedBuffer("Lower matrix");
|
|
|
|
}
|
|
|
|
for (int k = e * n2, l = 0; k < (e + 1) * n2; k++) {
|
|
|
|
output->p(k, lowerMatrix->e<T>(l++));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return ND4J_STATUS_OK;
|
|
|
|
}
|
|
|
|
|
|
|
|
int cholesky(nd4j::LaunchContext * context, NDArray* input, NDArray* output, bool inplace) {
|
2019-08-24 15:59:30 +02:00
|
|
|
BUILD_SINGLE_SELECTOR(input->dataType(), return cholesky_, (context, input, output, inplace), FLOAT_TYPES);
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
template <typename T>
|
2019-08-24 15:59:30 +02:00
|
|
|
int logdetFunctor_(LaunchContext *context, NDArray* input, NDArray* output) {
|
2019-12-20 20:35:39 +01:00
|
|
|
auto tempOutput = input->dup();
|
|
|
|
int res = cholesky_<T>(context, input, &tempOutput, false);
|
2019-06-06 14:21:15 +02:00
|
|
|
if (res != ND4J_STATUS_OK)
|
|
|
|
return res;
|
|
|
|
auto n = input->sizeAt(-1);
|
|
|
|
auto totalCount = output->lengthOf();
|
|
|
|
std::vector<T> d(n);
|
2019-12-20 20:35:39 +01:00
|
|
|
ResultSet matricies = tempOutput.allTensorsAlongDimension({input->rankOf()-2, input->rankOf() - 1});
|
2019-06-06 14:21:15 +02:00
|
|
|
|
2019-12-20 20:35:39 +01:00
|
|
|
for (Nd4jLong e = 0; e < totalCount; e++) {
|
|
|
|
for (size_t i = 0; i < n; ++i)
|
|
|
|
output->t<T>(e) += nd4j::math::nd4j_log<T,T>(nd4j::math::nd4j_pow<T,T,T>(matricies.at(e)->t<T>(i, i), T(2)));
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
return ND4J_STATUS_OK;
|
|
|
|
}
|
|
|
|
|
|
|
|
int logdetFunctor(nd4j::LaunchContext * context, NDArray* input, NDArray* output) {
|
2019-08-24 15:59:30 +02:00
|
|
|
BUILD_SINGLE_SELECTOR(input->dataType(), return logdetFunctor_, (context, input, output), FLOAT_TYPES);
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
|
2019-12-20 15:56:28 +01:00
|
|
|
int lup(nd4j::LaunchContext * context, NDArray* input, NDArray* compound, NDArray* permutation) {
|
|
|
|
BUILD_DOUBLE_SELECTOR(input->dataType(), permutation->dataType(), lup_, (context, input, compound, permutation), FLOAT_NATIVE, INDEXING_TYPES);
|
|
|
|
return Status::OK();
|
|
|
|
}
|
|
|
|
|
2019-06-06 14:21:15 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|