From c2514bc6b8b76a13d73f947ae5d7a67d4d7204c5 Mon Sep 17 00:00:00 2001
From: Tomas Oberhuber <tomas.oberhuber@fjfi.cvut.cz>
Date: Fri, 17 Jan 2020 17:40:18 +0100
Subject: [PATCH] Debuging assignment of tridiagonal matrix to sparse matrix.
---
src/TNL/Matrices/SparseMatrix.h | 2 +-
src/TNL/Matrices/SparseMatrix.hpp | 59 +++++++---
src/TNL/Matrices/TridiagonalMatrixRowView.hpp | 2 +-
src/TNL/Matrices/TridiagonalMatrixView.h | 2 +
src/TNL/Matrices/TridiagonalMatrixView.hpp | 17 +--
src/UnitTests/Matrices/SparseMatrixCopyTest.h | 105 +++++++++++++++++-
6 files changed, 155 insertions(+), 32 deletions(-)
diff --git a/src/TNL/Matrices/SparseMatrix.h b/src/TNL/Matrices/SparseMatrix.h
index 43ea25bf58..26d5d2d840 100644
--- a/src/TNL/Matrices/SparseMatrix.h
+++ b/src/TNL/Matrices/SparseMatrix.h
@@ -73,7 +73,7 @@ class SparseMatrix : public Matrix< Real, Device, Index, RealAllocator >
const RealAllocatorType& realAllocator = RealAllocatorType(),
const IndexAllocatorType& indexAllocator = IndexAllocatorType() );
- ViewType getView();
+ ViewType getView() const; // TODO: remove const
ConstViewType getConstView() const;
diff --git a/src/TNL/Matrices/SparseMatrix.hpp b/src/TNL/Matrices/SparseMatrix.hpp
index 14495ad3d5..6d3d9d8b8b 100644
--- a/src/TNL/Matrices/SparseMatrix.hpp
+++ b/src/TNL/Matrices/SparseMatrix.hpp
@@ -82,13 +82,13 @@ template< typename Real,
typename IndexAllocator >
auto
SparseMatrix< Real, Device, Index, MatrixType, Segments, RealAllocator, IndexAllocator >::
-getView() -> ViewType
+getView() const -> ViewType
{
return ViewType( this->getRows(),
this->getColumns(),
- this->getValues().getView(),
- this->columnIndexes.getView(),
- this->segments.getView() );
+ const_cast< SparseMatrix* >( this )->getValues().getView(), // TODO: remove const_cast
+ const_cast< SparseMatrix* >( this )->columnIndexes.getView(),
+ const_cast< SparseMatrix* >( this )->segments.getView() );
}
template< typename Real,
@@ -624,7 +624,8 @@ operator=( const Dense< Real_, Device_, Index_, RowMajorOrder, RealAllocator_ >&
thisValuesBuffer_view = matrixValuesBuffer_view;
////
- // Copy matrix elements from the buffer to the matrix
+ // Copy matrix elements from the buffer to the matrix and ignoring
+ // zero matrix elements.
const IndexType matrix_columns = this->getColumns();
auto f2 = [=] __cuda_callable__ ( IndexType rowIdx, IndexType localIdx, IndexType& columnIndex, RealType& value, bool& compute ) mutable {
RealType inValue( 0.0 );
@@ -672,34 +673,41 @@ operator=( const RHSMatrix& matrix )
using RHSRealType = typename RHSMatrix::RealType;
using RHSDeviceType = typename RHSMatrix::DeviceType;
using RHSRealAllocatorType = typename RHSMatrix::RealAllocatorType;
- using RHSIndexAllocatorType = typename RHSMatrix::IndexAllocatorType;
+ using RHSIndexAllocatorType = typename Allocators::Default< RHSDeviceType >::template Allocator< RHSIndexType >;
- typename RHSMatrix::RowsCapacitiesType rowLengths;
+ Containers::Vector< RHSIndexType, RHSDeviceType, RHSIndexType > rowLengths;
matrix.getCompressedRowLengths( rowLengths );
this->setDimensions( matrix.getRows(), matrix.getColumns() );
this->setCompressedRowLengths( rowLengths );
+ Containers::Vector< IndexType, DeviceType, IndexType > rowLocalIndexes( matrix.getRows() );
+ rowLocalIndexes = 0;
+
// TODO: use getConstView when it works
const auto matrixView = const_cast< RHSMatrix& >( matrix ).getView();
const IndexType paddingIndex = this->getPaddingIndex();
auto columns_view = this->columnIndexes.getView();
auto values_view = this->values.getView();
+ auto rowLocalIndexes_view = rowLocalIndexes.getView();
columns_view = paddingIndex;
- /*if( std::is_same< DeviceType, RHSDeviceType >::value )
+ if( std::is_same< DeviceType, RHSDeviceType >::value )
{
const auto segments_view = this->segments.getView();
- auto f = [=] __cuda_callable__ ( RHSIndexType rowIdx, RHSIndexType localIdx, RHSIndexType columnIndex, const RHSRealType& value, bool& compute ) mutable {
+ auto f = [=] __cuda_callable__ ( RHSIndexType rowIdx, RHSIndexType localIdx_, RHSIndexType columnIndex, const RHSRealType& value, bool& compute ) mutable {
+ RealType inValue( 0.0 );
+ IndexType localIdx( rowLocalIndexes_view[ rowIdx ] );
if( columnIndex != paddingIndex )
{
- IndexType thisGlobalIdx = segments_view.getGlobalIndex( rowIdx, localIdx );
+ IndexType thisGlobalIdx = segments_view.getGlobalIndex( rowIdx, localIdx++ );
columns_view[ thisGlobalIdx ] = columnIndex;
values_view[ thisGlobalIdx ] = value;
+ rowLocalIndexes_view[ rowIdx ] = localIdx;
}
};
matrix.forAllRows( f );
}
- else*/
+ else
{
const IndexType maxRowLength = max( rowLengths );
const IndexType bufferRowsCount( 128 );
@@ -739,14 +747,29 @@ operator=( const RHSMatrix& matrix )
thisColumnsBuffer_view = matrixColumnsBuffer_view;
////
- // Copy matrix elements from the buffer to the matrix
- auto f2 = [=] __cuda_callable__ ( IndexType rowIdx, IndexType localIdx, IndexType& columnIndex, RealType& value, bool& compute ) mutable {
- const IndexType bufferIdx = ( rowIdx - baseRow ) * maxRowLength + localIdx;
- const IndexType column = thisColumnsBuffer_view[ bufferIdx ];
- if( column != paddingIndex )
+ // Copy matrix elements from the buffer to the matrix and ignoring
+ // zero matrix elements
+ const IndexType matrix_columns = this->getColumns();
+ auto matrix_view = matrix.getView();
+ auto f2 = [=] __cuda_callable__ ( IndexType rowIdx, IndexType localIdx_, IndexType& columnIndex, RealType& value, bool& compute ) mutable {
+ RealType inValue( 0.0 );
+ IndexType bufferIdx, localIdx( rowLocalIndexes_view[ rowIdx ] );
+ auto matrixRow = matrix_view.getRow( rowIdx );
+ while( inValue == 0.0 && localIdx < matrixRow.getSize() ) //matrix_columns )
{
- columnIndex = column;
- value = thisValuesBuffer_view[ bufferIdx ];
+ bufferIdx = ( rowIdx - baseRow ) * maxRowLength + localIdx++;
+ inValue = thisValuesBuffer_view[ bufferIdx ];
+ }
+ rowLocalIndexes_view[ rowIdx ] = localIdx;
+ if( inValue == 0.0 )
+ {
+ columnIndex = paddingIndex;
+ value = 0.0;
+ }
+ else
+ {
+ columnIndex = thisColumnsBuffer_view[ bufferIdx ];//column - 1;
+ value = inValue;
}
};
this->forRows( baseRow, lastRow, f2 );
diff --git a/src/TNL/Matrices/TridiagonalMatrixRowView.hpp b/src/TNL/Matrices/TridiagonalMatrixRowView.hpp
index ba60876b9b..80fc1a26d5 100644
--- a/src/TNL/Matrices/TridiagonalMatrixRowView.hpp
+++ b/src/TNL/Matrices/TridiagonalMatrixRowView.hpp
@@ -29,7 +29,7 @@ auto
TridiagonalMatrixRowView< ValuesView, Indexer >::
getSize() const -> IndexType
{
- return indexer.getRowSize();
+ return indexer.getRowSize( rowIdx );
}
template< typename ValuesView, typename Indexer >
diff --git a/src/TNL/Matrices/TridiagonalMatrixView.h b/src/TNL/Matrices/TridiagonalMatrixView.h
index 7db517cbd3..61b005c5a7 100644
--- a/src/TNL/Matrices/TridiagonalMatrixView.h
+++ b/src/TNL/Matrices/TridiagonalMatrixView.h
@@ -75,8 +75,10 @@ class TridiagonalMatrixView : public MatrixView< Real, Device, Index >
template< typename Real_, typename Device_, typename Index_, bool RowMajorOrder_ >
bool operator != ( const TridiagonalMatrixView< Real_, Device_, Index_, RowMajorOrder_ >& matrix ) const;
+ __cuda_callable__
RowView getRow( const IndexType& rowIdx );
+ __cuda_callable__
const RowView getRow( const IndexType& rowIdx ) const;
void setValue( const RealType& v );
diff --git a/src/TNL/Matrices/TridiagonalMatrixView.hpp b/src/TNL/Matrices/TridiagonalMatrixView.hpp
index d8fa6061c3..7fc5fd6b7b 100644
--- a/src/TNL/Matrices/TridiagonalMatrixView.hpp
+++ b/src/TNL/Matrices/TridiagonalMatrixView.hpp
@@ -340,25 +340,26 @@ forRows( IndexType first, IndexType last, Function& function ) const
{
const auto values_view = this->values.getConstView();
const auto indexer = this->indexer;
+ bool compute( true );
auto f = [=] __cuda_callable__ ( IndexType rowIdx ) mutable {
if( rowIdx == 0 )
{
- function( 0, 0, 0, values_view[ indexer.getGlobalIndex( 0, 0 ) ] );
- function( 0, 1, 1, values_view[ indexer.getGlobalIndex( 0, 1 ) ] );
+ function( 0, 0, 0, values_view[ indexer.getGlobalIndex( 0, 0 ) ], compute );
+ function( 0, 1, 1, values_view[ indexer.getGlobalIndex( 0, 1 ) ], compute );
}
else if( rowIdx + 1 < indexer.getColumns() )
{
- function( rowIdx, 0, rowIdx - 1, values_view[ indexer.getGlobalIndex( rowIdx, 0 ) ] );
- function( rowIdx, 1, rowIdx, values_view[ indexer.getGlobalIndex( rowIdx, 1 ) ] );
- function( rowIdx, 2, rowIdx + 1, values_view[ indexer.getGlobalIndex( rowIdx, 2 ) ] );
+ function( rowIdx, 0, rowIdx - 1, values_view[ indexer.getGlobalIndex( rowIdx, 0 ) ], compute );
+ function( rowIdx, 1, rowIdx, values_view[ indexer.getGlobalIndex( rowIdx, 1 ) ], compute );
+ function( rowIdx, 2, rowIdx + 1, values_view[ indexer.getGlobalIndex( rowIdx, 2 ) ], compute );
}
else if( rowIdx < indexer.getColumns() )
{
- function( rowIdx, 0, rowIdx - 1, values_view[ indexer.getGlobalIndex( rowIdx, 0 ) ] );
- function( rowIdx, 1, rowIdx, values_view[ indexer.getGlobalIndex( rowIdx, 1 ) ] );
+ function( rowIdx, 0, rowIdx - 1, values_view[ indexer.getGlobalIndex( rowIdx, 0 ) ], compute );
+ function( rowIdx, 1, rowIdx, values_view[ indexer.getGlobalIndex( rowIdx, 1 ) ], compute );
}
else
- function( rowIdx, 0, rowIdx, values_view[ indexer.getGlobalIndex( rowIdx, 0 ) ] );
+ function( rowIdx, 0, rowIdx, values_view[ indexer.getGlobalIndex( rowIdx, 0 ) ], compute );
};
Algorithms::ParallelFor< DeviceType >::exec( first, last, f );
}
diff --git a/src/UnitTests/Matrices/SparseMatrixCopyTest.h b/src/UnitTests/Matrices/SparseMatrixCopyTest.h
index e9898bb393..b285f5e058 100644
--- a/src/UnitTests/Matrices/SparseMatrixCopyTest.h
+++ b/src/UnitTests/Matrices/SparseMatrixCopyTest.h
@@ -15,6 +15,8 @@
#include <TNL/Matrices/SparseMatrix.h>
#include <TNL/Matrices/MatrixType.h>
#include <TNL/Matrices/Dense.h>
+#include <TNL/Matrices/Tridiagonal.h>
+#include <TNL/Matrices/Multidiagonal.h>
#include <TNL/Containers/Segments/CSR.h>
#include <TNL/Containers/Segments/Ellpack.h>
#include <TNL/Containers/Segments/SlicedEllpack.h>
@@ -370,6 +372,7 @@ void testCopyAssignment()
Matrix2 triDiag2;
triDiag2 = triDiag1;
+ checkTriDiagMatrix( triDiag1 );
checkTriDiagMatrix( triDiag2 );
}
{
@@ -390,6 +393,7 @@ void testCopyAssignment()
Matrix2 unevenRowSize2;
unevenRowSize2 = unevenRowSize1;
+
checkUnevenRowSizeMatrix( unevenRowSize2 );
}
}
@@ -435,6 +439,62 @@ void testConversion()
}
}
+template< typename Matrix >
+void tridiagonalMatrixAssignment()
+{
+ using RealType = typename Matrix::RealType;
+ using DeviceType = typename Matrix::DeviceType;
+ using IndexType = typename Matrix::IndexType;
+
+ using TridiagonalHost = TNL::Matrices::Tridiagonal< RealType, TNL::Devices::Host, IndexType >;
+ using TridiagonalCuda = TNL::Matrices::Tridiagonal< RealType, TNL::Devices::Cuda, IndexType >;
+
+ const IndexType rows( 10 ), columns( 10 );
+ TridiagonalHost hostMatrix( rows, columns );
+ for( IndexType i = 0; i < columns; i++ )
+ for( IndexType j = TNL::max( 0, i - 1 ); j < TNL::min( rows, i + 2 ); j++ )
+ hostMatrix.setElement( i, j, i + j );
+
+ std::cerr << hostMatrix << std::endl;
+ Matrix matrix;
+ matrix = hostMatrix;
+ std::cerr << matrix << std::endl;
+ using RowCapacitiesType = typename Matrix::RowsCapacitiesType;
+ RowCapacitiesType rowCapacities;
+ matrix.getCompressedRowLengths( rowCapacities );
+ RowCapacitiesType exactRowLengths{ 0, 3, 3, 3, 3, 3, 3, 3, 3, 2 };
+ EXPECT_EQ( rowCapacities, exactRowLengths );
+ for( IndexType i = 0; i < columns; i++ )
+ for( IndexType j = 0; j < rows; j++ )
+ {
+ if( abs( i - j ) > 1 )
+ EXPECT_EQ( matrix.getElement( i, j ), 0.0 );
+ else
+ EXPECT_EQ( matrix.getElement( i, j ), i + j );
+ }
+
+#ifdef HAVE_CUDA
+ TridiagonalCuda cudaMatrix( rows, columns );
+ cudaMatrix = hostMatrix;
+ /*for( IndexType i = 0; i < columns; i++ )
+ for( IndexType j = TNL::max( 0, i - 1 ); j < TNL::min( row, i + 1 ); j++ )
+ cudaMatrix.setElement( i, j, i + j );*/
+
+ matrix = cudaMatrix;
+ matrix.getCompressedRowLengths( rowCapacities );
+ EXPECT_EQ( rowCapacities, exactRowLengths );
+ for( IndexType i = 0; i < columns; i++ )
+ for( IndexType j = 0; j < rows; j++ )
+ {
+ if( abs( i - j ) > 1 )
+ EXPECT_EQ( matrix.getElement( i, j ), 0.0 );
+ else
+ EXPECT_EQ( matrix.getElement( i, j ), i + j );
+ }
+#endif
+
+}
+
template< typename Matrix >
void denseMatrixAssignment()
{
@@ -469,10 +529,10 @@ void denseMatrixAssignment()
#ifdef HAVE_CUDA
DenseCuda cudaMatrix( rows, columns );
- //cudaMatrix = hostMatrix;
- for( IndexType i = 0; i < columns; i++ )
+ cudaMatrix = hostMatrix;
+ /*for( IndexType i = 0; i < columns; i++ )
for( IndexType j = 0; j <= i; j++ )
- cudaMatrix.setElement( i, j, i + j );
+ cudaMatrix.setElement( i, j, i + j );*/
matrix = cudaMatrix;
matrix.getCompressedRowLengths( rowCapacities );
@@ -487,7 +547,7 @@ void denseMatrixAssignment()
}
#endif
}
-
+/*
TEST( SparseMatrixCopyTest, CSR_HostToHost )
{
testCopyAssignment< CSR_host, CSR_host >();
@@ -619,6 +679,43 @@ TEST( SparseMatrixCopyTest, SlicedEllpack_to_Ellpack_cuda )
testConversion< SE_cuda, E_cuda >();
}
#endif
+*/
+
+////
+// Tridiagonal matrix assignment test
+TEST( SparseMatrixCopyTest, TridiagonalMatrixAssignment_to_CSR_host )
+{
+ tridiagonalMatrixAssignment< CSR_host >();
+}
+
+TEST( SparseMatrixCopyTest, TridiagonalMatrixAssignment_to_Ellpack_host )
+{
+ tridiagonalMatrixAssignment< E_host >();
+}
+
+TEST( SparseMatrixCopyTest, TridiagonalMatrixAssignment_to_SlicedEllpack_host )
+{
+ tridiagonalMatrixAssignment< SE_host >();
+}
+
+#ifdef HAVE_CUDA
+TEST( SparseMatrixCopyTest, TridiagonalMatrixAssignment_to_CSR_cuda )
+{
+ tridiagonalMatrixAssignment< CSR_cuda >();
+}
+
+TEST( SparseMatrixCopyTest, TridiagonalMatrixAssignment_to_Ellpack_cuda )
+{
+ tridiagonalMatrixAssignment< E_cuda >();
+}
+
+TEST( SparseMatrixCopyTest, TridiagonalMatrixAssignment_to_SlicedEllpack_cuda )
+{
+ tridiagonalMatrixAssignment< SE_cuda >();
+}
+#endif // HAVE_CUDA
+
+
// Dense matrix assignment test
TEST( SparseMatrixCopyTest, DenseMatrixAssignment_to_CSR_host )
--
GitLab