cavis/libnd4j/include/ops/impl/specials_single.hpp

642 lines
21 KiB
C++

/*******************************************************************************
* 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, created on 07.10.2017.
// @author Yurii Shyrma (iuriish@yahoo.com)
//
#include <system/pointercast.h>
#include <helpers/shape.h>
#include <helpers/TAD.h>
#include <ops/specials.h>
#include <system/dll.h>
#include <array/NDArray.h>
#include <ops/declarable/CustomOperations.h>
#include <types/types.h>
#include <helpers/Loops.h>
namespace sd {
/**
* Concatneate multi array of the same shape together
* along a particular dimension
*/
// template <typename T>
// void SpecialMethods<T>::concatCpuGeneric(const std::vector<NDArray*>& inArrs, NDArray& output, const int axis) {
// const uint numOfArrs = inArrs.size();
// int outDim;
// const bool isOutputVector = output.isCommonVector(outDim);
// if(isOutputVector || (axis == 0 && output.ordering() == 'c')) {
// bool allVectorsOrScalars = true;
// const uint outEws = isOutputVector ? output.stridesOf()[outDim] : output.ews();
// std::vector<int> nonUnityDim(numOfArrs);
// std::vector<Nd4jLong> zOffset(numOfArrs);
// for(int i = 0; i < numOfArrs; i++) {
// allVectorsOrScalars &= (inArrs[i]->lengthOf() == 1 || inArrs[i]->isCommonVector(nonUnityDim[i]));
// if(!allVectorsOrScalars)
// break;
// if(i == 0) zOffset[0] = 0;
// else zOffset[i] = zOffset[i - 1] + outEws * inArrs[i - 1]->lengthOf();
// }
// if(allVectorsOrScalars) {
// T* outBuff = output.bufferAsT<T>();
// auto func = PRAGMA_THREADS_FOR {
// for (auto r = start; r < stop; r += increment) {
// const Nd4jLong arrLen = inArrs[r]->lengthOf();
// const uint xEws = (arrLen == 1) ? 1 : inArrs[r]->stridesOf()[nonUnityDim[r]];
// T *z = outBuff + zOffset[r];
// T *x = inArrs[r]->bufferAsT<T>();
// if (outEws == 1 && xEws == 1)
// for (Nd4jLong e = 0; e < arrLen; e++)
// z[e] = x[e];
// else
// for (Nd4jLong e = 0; e < arrLen; e++)
// z[e * outEws] = x[e * xEws];
// }
// };
// samediff::Threads::parallel_tad(func, 0, numOfArrs);
// return;
// }
// }
// const int rank = inArrs[0]->rankOf();
// const int rank2 = 2*rank;
// std::vector<std::vector<Nd4jLong>> indices(numOfArrs, std::vector<Nd4jLong>(rank2,0));
// // take into account indices for first array
// indices[0][2 * axis + 1] = inArrs[0]->sizeAt(axis);
// // loop through the rest of input arrays
// for(int i = 1; i < numOfArrs; ++i) {
// indices[i][2 * axis] = indices[i-1][2 * axis + 1]; // index start from
// indices[i][2 * axis + 1] = indices[i-1][2 * axis + 1] + inArrs[i]->sizeAt(axis); // index end with (excluding)
// }
// auto func = PRAGMA_THREADS_FOR {
// for (auto i = start; i < stop; i += increment) {
// auto temp = output(indices[i], true);
// sd::TransformLoops<T, T, T>::template loopTransform<simdOps::Assign<T, T>>( inArrs[i]->bufferAsT<T>(), inArrs[i]->shapeInfo(), temp.bufferAsT<T>(), temp.shapeInfo(), nullptr, 0, 1);
// }
// };
// samediff::Threads::parallel_tad(func, 0, numOfArrs);
// }
template <typename T>
void SpecialMethods<T>::concatCpuGeneric(const std::vector<const NDArray*>& inArrs, NDArray& output, const int axis) {
const int numOfInArrs = inArrs.size();
const auto sizeofT = output.sizeOfT();
T* zBuff = output.bufferAsT<T>();
bool luckCase1 = ((axis == 0 && output.ordering() == 'c') || (axis == output.rankOf() - 1 && output.ordering() == 'f')) && output.ews() == 1;
if(luckCase1) {
for (uint i = 0; i < numOfInArrs; ++i) {
luckCase1 &= inArrs[i]->ordering() == output.ordering() && inArrs[i]->ews() == 1;
if(!luckCase1)
break;
}
}
if(luckCase1) { // for example {1,10} + {2,10} + {3,10} = {6, 10} order c; or {10,1} + {10,2} + {10,3} = {10, 6} order f
T* z = zBuff;
for (uint i = 0; i < numOfInArrs; ++i) {
const auto memAmountToCopy = inArrs[i]->lengthOf();
memcpy(z, inArrs[i]->bufferAsT<T>(), memAmountToCopy * sizeofT);
z += memAmountToCopy;
}
return;
}
// const bool isZcontin = output.strideAt(axis) == 1;
// bool areInputsContin = true;
// bool allSameOrder = true;
// std::vector<Nd4jLong> strideOfContigStride(numOfInArrs);
// if(isZcontin) {
// for (uint i = 0; i < numOfInArrs; ++i) {
// areInputsContin &= inArrs[i]->strideAt(axis) == 1;
// allSameOrder &= inArrs[i]->ordering() == output.ordering();
// if(!areInputsContin || !allSameOrder)
// break;
// strideOfContigStride[i] = shape::strideOverContigAxis(axis, inArrs[i]->shapeInfo());
// }
// }
// const bool luckCase2 = isZcontin && areInputsContin && allSameOrder;
// if(luckCase2) { // for example {2,1,3} + {2,5,3} + {2,10,3} = {2,16,3}, here axis 1 shoud have stride = 1 for all inputs arrays and output array
// const auto zStep = shape::strideOverContigAxis(axis, output.shapeInfo());
// for (uint i = 0; i < output.lengthOf() / output.sizeAt(axis); ++i) {
// T* z = zBuff + zStep * i;
// for (uint j = 0; j < inArrs.size(); ++j) {
// const auto xDim = inArrs[j]->sizeAt(axis);
// const T* x = inArrs[j]->bufferAsT<T>() + strideOfContigStride[j] * i;
// memcpy(z, x, xDim * sizeofT);
// z += xDim;
// }
// }
// return;
// }
// general case
auto func = PRAGMA_THREADS_FOR {
int coords[MAX_RANK], temp;
for (auto i = start; i < stop; i += increment) {
shape::index2coordsCPU(start, i, output.shapeInfo(), coords);
const auto zOffset = shape::getOffset(output.shapeInfo(), coords);
uint inArrIdx = 0;
uint xDim = inArrs[inArrIdx]->sizeAt(axis);
temp = coords[axis];
while (coords[axis] >= xDim) {
coords[axis] -= xDim;
xDim = inArrs[++inArrIdx]->sizeAt(axis);
}
const T* x = inArrs[inArrIdx]->bufferAsT<T>();
const auto xOffset = shape::getOffset(inArrs[inArrIdx]->shapeInfo(), coords);
zBuff[zOffset] = x[xOffset];
coords[axis] = temp;
}
};
samediff::Threads::parallel_for(func, 0, output.lengthOf());
}
/**
* Concatneate multi array of the same shape together
* along a particular dimension
*/
template <typename T>
void SpecialMethods<T>::concatCpuGeneric(int dimension, int numArrays, Nd4jPointer *data, Nd4jPointer *inputShapeInfo, void *vresult, Nd4jLong const* resultShapeInfo) {
auto result = reinterpret_cast<T *>(vresult);
std::vector<const NDArray*> inputs(numArrays);
NDArray output(static_cast<void*>(result), resultShapeInfo);
for(int i = 0; i < numArrays; ++i)
inputs[i] = new NDArray(static_cast<void *>(data[i]), static_cast<Nd4jLong*>(inputShapeInfo[i]));
sd::SpecialMethods<T>::concatCpuGeneric(inputs, output, dimension);
for(int i = 0; i < numArrays; ++i)
delete inputs[i];
}
template <typename T>
void SpecialMethods<T>::splitCpuGeneric(const NDArray& input, const std::vector<NDArray*>& outArrs, const int axis) {
int numSplits = outArrs.size();
const auto sizeofT = input.sizeOfT();
auto xBuff = input.bufferAsT<T>();
bool luckCase1 = ((axis == 0 && input.ordering() == 'c') || (axis == input.rankOf() - 1 && input.ordering() == 'f')) && input.ews() == 1;
if (luckCase1) {
for (uint i = 0; i < numSplits; ++i) {
luckCase1 &= outArrs[i]->ordering() == input.ordering() && outArrs[i]->ews() == 1;
if (!luckCase1)
break;
}
}
if (luckCase1) {
T* x = const_cast<T*>(xBuff);
for (uint i = 0; i < numSplits; ++i) {
const auto memAmountToCopy = outArrs[i]->lengthOf();
memcpy(outArrs[i]->bufferAsT<T>(), x, memAmountToCopy * sizeofT);
x += memAmountToCopy;
}
return;
}
// const bool isXcontin = input.strideAt(axis) == 1;
// bool areOutsContin = true;
// bool allSameOrder = true;
// std::vector<Nd4jLong> strideOfContigStride(numSplits);
// if (isXcontin) {
// for (uint i = 0; i < numSplits; ++i) {
// areOutsContin &= outArrs[i]->strideAt(axis) == 1;
// allSameOrder &= outArrs[i]->ordering() == input.ordering();
// if (!areOutsContin || !allSameOrder)
// break;
// strideOfContigStride[i] = shape::strideOverContigAxis(axis, outArrs[i]->shapeInfo());
// }
// }
// const bool luckCase2 = isXcontin && areOutsContin && allSameOrder;
// if (luckCase2) {
// const auto xStep = shape::strideOverContigAxis(axis, input.shapeInfo());
// for (uint i = 0; i < input.lengthOf() / input.sizeAt(axis); ++i) {
// T* x = xBuff + xStep * i;
// for (uint j = 0; j < numSplits; ++j) {
// const auto zDim = outArrs[j]->sizeAt(axis);
// T* z = outArrs[j]->bufferAsT<T>() + strideOfContigStride[j] * i;
// memcpy(z, x, zDim * sizeofT);
// x += zDim;
// }
// }
// return;
// }
uint zDim = outArrs[0]->sizeAt(axis);
// general case
auto func = PRAGMA_THREADS_FOR{
int coords[MAX_RANK], temp;
for (auto i = start; i < stop; i += increment) {
shape::index2coordsCPU(start, i, input.shapeInfo(), coords);
const auto xOffset = shape::getOffset(input.shapeInfo(), coords);
uint outArrIdx = 0;
temp = coords[axis];
while (coords[axis] >= zDim) {
coords[axis] -= zDim;
++outArrIdx;
}
T* z = outArrs[outArrIdx]->bufferAsT<T>();
const auto zOffset = shape::getOffset(outArrs[outArrIdx]->shapeInfo(), coords);
z[zOffset] = xBuff[xOffset];
coords[axis] = temp;
}
};
samediff::Threads::parallel_for(func, 0, input.lengthOf());
}
/**
* This kernel accumulates X arrays, and stores result into Z
*
* @tparam T
* @param x
* @param z
* @param n
* @param length
*/
template<typename T>
void SpecialMethods<T>::accumulateGeneric(void **vx, void *vz, Nd4jLong const* zShapeInfo, int n, const Nd4jLong length) {
auto z = reinterpret_cast<T *>(vz);
auto x = reinterpret_cast<T **>(vx);
auto func = PRAGMA_THREADS_FOR {
for (auto i = start; i < stop; i++) {
for (auto ar = 0L; ar < n; ar++) {
z[i] += x[ar][i];
}
}
};
samediff::Threads::parallel_for(func, 0, length);
}
/**
* This kernel averages X input arrays, and stores result to Z
*
* @tparam T
* @param x
* @param z
* @param n
* @param length
* @param propagate
*/
template<typename T>
void SpecialMethods<T>::averageGeneric(void **vx, void *vz, Nd4jLong const* zShapeInfo, int n, const Nd4jLong length, bool propagate) {
auto z = reinterpret_cast<T *>(vz);
auto x = reinterpret_cast<T **>(vx);
if (z == nullptr) {
//code branch for absent Z
z = x[0];
PRAGMA_OMP_SIMD
for (uint64_t i = 0; i < length; i++) {
z[i] /= static_cast<T>(n);
}
auto func = PRAGMA_THREADS_FOR {
for (auto i = start; i < stop; i++) {
for (Nd4jLong ar = 1; ar < n; ar++) {
z[i] += x[ar][i] / static_cast<T>(n);
}
}
};
samediff::Threads::parallel_for(func, 0, length);
// instead of doing element-wise propagation, we just issue memcpy to propagate data
for (Nd4jLong ar = 1; ar < n; ar++) {
memcpy(x[ar], z, length * sizeof(T));
}
} else {
// code branch for existing Z
// memset before propagation
memset(z, 0, length * sizeof(T));
// aggregation step
auto func = PRAGMA_THREADS_FOR {
for (auto i = start; i < stop; i++) {
for (Nd4jLong ar = 0; ar < n; ar++) {
z[i] += x[ar][i] / static_cast<T>(n);
}
}
};
samediff::Threads::parallel_for(func, 0, length);
// instead of doing element-wise propagation, we just issue memcpy to propagate data
for (Nd4jLong ar = 0; ar < n; ar++) {
memcpy(x[ar], z, length * sizeof(T));
}
}
}
template <typename T>
Nd4jLong SpecialMethods<T>::getPosition(Nd4jLong const* xShapeInfo, Nd4jLong index) {
auto xEWS = shape::elementWiseStride(xShapeInfo);
if (xEWS == 1)
return index;
else if (xEWS > 1)
return index * xEWS;
else
return shape::getIndexOffset(index, xShapeInfo);
}
template<typename T>
void SpecialMethods<T>::quickSort_parallel_internal(T* array, Nd4jLong const* xShapeInfo, int left, int right, int cutoff, bool descending) {
int i = left, j = right;
T tmp;
T pivot = array[getPosition(xShapeInfo, (left + right) / 2)];
{
/* PARTITION PART */
while (i <= j) {
if (descending) {
while (array[getPosition(xShapeInfo, i)] > pivot)
i++;
while (array[getPosition(xShapeInfo, j)] < pivot)
j--;
if (i <= j) {
tmp = array[getPosition(xShapeInfo, i)];
array[getPosition(xShapeInfo, i)] = array[getPosition(xShapeInfo, j)];
array[getPosition(xShapeInfo, j)] = tmp;
i++;
j--;
}
} else {
while (array[getPosition(xShapeInfo, i)] < pivot)
i++;
while (array[getPosition(xShapeInfo, j)] > pivot)
j--;
if (i <= j) {
tmp = array[getPosition(xShapeInfo, i)];
array[getPosition(xShapeInfo, i)] = array[getPosition(xShapeInfo, j)];
array[getPosition(xShapeInfo, j)] = tmp;
i++;
j--;
}
}
}
}
//
if ( ((right-left)<cutoff) ){
if (left < j){ quickSort_parallel_internal(array, xShapeInfo, left, j, cutoff, descending); }
if (i < right){ quickSort_parallel_internal(array, xShapeInfo, i, right, cutoff, descending); }
}else{
PRAGMA_OMP_TASK
{ quickSort_parallel_internal(array, xShapeInfo, left, j, cutoff, descending); }
PRAGMA_OMP_TASK
{ quickSort_parallel_internal(array, xShapeInfo, i, right, cutoff, descending); }
}
}
template<typename T>
void SpecialMethods<T>::quickSort_parallel(void *varray, Nd4jLong const* xShapeInfo, Nd4jLong lenArray, int numThreads, bool descending){
auto array = reinterpret_cast<T *>(varray);
int cutoff = 1000;
PRAGMA_OMP_PARALLEL_THREADS(numThreads)
{
PRAGMA_OMP_SINGLE_ARGS(nowait)
{
quickSort_parallel_internal(array, xShapeInfo, 0, lenArray-1, cutoff, descending);
}
}
}
template <typename T>
int SpecialMethods<T>::nextPowerOf2(int number) {
int pos = 0;
while (number > 0) {
pos++;
number = number >> 1;
}
return (int) pow(2, pos);
}
template <typename T>
int SpecialMethods<T>::lastPowerOf2(int number) {
int p = 1;
while (p <= number)
p <<= 1;
p >>= 1;
return p;
}
template<typename T>
void SpecialMethods<T>::sortGeneric(void *vx, Nd4jLong const* xShapeInfo, bool descending) {
auto x = reinterpret_cast<T *>(vx);
quickSort_parallel(x, xShapeInfo, shape::length(xShapeInfo), omp_get_max_threads(), descending);
}
template<typename T>
void SpecialMethods<T>::sortTadGeneric(void *vx, Nd4jLong const* xShapeInfo, int *dimension, int dimensionLength, Nd4jLong const* tadShapeInfo, Nd4jLong const* tadOffsets, bool descending) {
auto x = reinterpret_cast<T *>(vx);
//quickSort_parallel(x, xShapeInfo, shape::length(xShapeInfo), omp_get_max_threads(), descending);
Nd4jLong xLength = shape::length(xShapeInfo);
Nd4jLong xTadLength = shape::tadLength(xShapeInfo, dimension, dimensionLength);
int numTads = xLength / xTadLength;
auto func = PRAGMA_THREADS_FOR {
for (auto r = start; r < stop; r++) {
T *dx = x + tadOffsets[r];
quickSort_parallel(dx, tadShapeInfo, xTadLength, 1, descending);
}
};
samediff::Threads::parallel_tad(func, 0, numTads);
}
template<typename T>
void SpecialMethods<T>::decodeBitmapGeneric(const void *dx, Nd4jLong N, void *vz, Nd4jLong const* zShapeInfo) {
auto dz = reinterpret_cast<T *>(vz);
auto x = reinterpret_cast<const int *>(dx);
Nd4jLong lim = N / 16 + 5;
FloatBits2 fb;
fb.i_ = x[2];
float threshold = fb.f_;
auto pPos = -1;
auto func = PRAGMA_THREADS_FOR {
for (auto e = start; e < stop; e++) {
const auto v = x[e];
for (int bitId = 0; bitId < 16; bitId++) {
bool hasBit = (v & 1 << (bitId)) != 0;
bool hasSign = (v & 1 << (bitId + 16)) != 0;
auto cPos = (e - 4) * 16 + bitId;
if (hasBit) {
if (hasSign)
dz[cPos] -= static_cast<T>(threshold);
else
dz[cPos] += static_cast<T>(threshold);
} else if (hasSign) {
dz[cPos] -= static_cast<T>(threshold / 2);
}
pPos = cPos;
}
}
};
samediff::Threads::parallel_for(func, 4, lim);
}
template<typename T>
Nd4jLong SpecialMethods<T>::encodeBitmapGeneric(void *vx, Nd4jLong const* xShapeInfo, Nd4jLong N, int *dz, float threshold) {
auto dx = reinterpret_cast<T *>(vx);
const T two(2.0f);
const T zero(0.0f);
const T t(threshold);
const T thalf = t / two;
//auto func = PRAGMA_REDUCE_LONG {
Nd4jLong retVal = 0L;
PRAGMA_OMP_PARALLEL_FOR_REDUCTION(+:retVal)
for (auto x = 0; x < N; x += 16) {
int byte = 0;
int byteId = x / 16 + 4;
for (int f = 0; f < 16; f++) {
auto e = x + f;
if (e >= N)
continue;
T val = dx[e];
T abs = sd::math::nd4j_abs<T>(val);
int bitId = e % 16;
if (abs >= t) {
byte |= 1 << (bitId);
retVal++;
if (val < zero) {
byte |= 1 << (bitId + 16);
dx[e] += t;
} else {
dx[e] -= t;
}
} else if (abs >= thalf && val < zero) {
byte |= 1 << (bitId + 16);
dx[e] += thalf;
retVal++;
}
}
dz[byteId] = byte;
}
return retVal;
//};
//return samediff::Threads::parallel_long(func, LAMBDA_SUML, 0, N, 16);
}
}