File: polynomial.inl

package info (click to toggle)
python-escript 5.0-3
  • links: PTS, VCS
  • area: main
  • in suites: stretch
  • size: 87,772 kB
  • ctags: 49,550
  • sloc: python: 585,488; cpp: 133,173; ansic: 18,675; xml: 3,283; sh: 690; makefile: 215
file content (213 lines) | stat: -rw-r--r-- 6,684 bytes parent folder | download | duplicates (4)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
/*
 *  Copyright 2008-2009 NVIDIA Corporation
 *
 *  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.
 */

/*! \file polynomial.inl
 *  \brief Inline file for polynomial.h
 */

#include <cusp/multiply.h>

#include <cusp/detail/format_utils.h>
#include <cusp/detail/spectral_radius.h>

#include <math.h>

namespace cusp
{
namespace relaxation
{
namespace detail
{
template <typename ValueType>
void chebyshev_polynomial_coefficients( const ValueType rho,
                                        cusp::array1d<ValueType,cusp::host_memory>& coefficients,
                                        const ValueType lower_bound = 1.0/30.0,
                                        const ValueType upper_bound = 1.1)
{
    const size_t degree = 3;

    ValueType x0 = lower_bound * rho;
    ValueType x1 = upper_bound * rho;

    // Chebyshev roots for the interval [-1,1]
    cusp::array1d<ValueType,cusp::host_memory> std_roots(degree);

    for( size_t i=0; i<degree; i++ )
        std_roots[i] = std::cos( M_PI * (ValueType(i) + 0.5)/ degree );

    // Chebyshev roots for the interval [x0,x1]
    for( size_t i=0; i<degree; i++ )
        std_roots[i] = 0.5 * (x1-x0) * (1 + std_roots[i]) + x0;

    // Compute monic polynomial coefficients of polynomial with scaled roots
    // TODO: Implement convolution method for polynomial multiplication
    coefficients.resize(degree+1);
    ValueType a = std_roots[0];
    ValueType b = std_roots[1];
    ValueType c = std_roots[2];
    coefficients[0] = 1.0;
    coefficients[1] = -(a+b+c);
    coefficients[2] = (a*b) + (b*c) + (c*a);
    coefficients[3] = -(a*b*c);

    // Scale coefficients to enforce C(0) = 1.0
    ValueType scale_factor = 1.0/coefficients.back();
    cusp::blas::scal(coefficients, scale_factor);
}
}

// constructor
template <typename ValueType, typename MemorySpace>
polynomial<ValueType,MemorySpace>
::polynomial()
{
}

template <typename ValueType, typename MemorySpace>
template<typename MatrixType, typename VectorType>
polynomial<ValueType,MemorySpace>
::polynomial(const MatrixType& A, const VectorType& coefficients)
{
    size_t default_size = coefficients.size()-1;
    default_coefficients.resize( default_size );
    for( size_t index = 0; index < default_size; index++ )
        default_coefficients[index] = -coefficients[index];

    size_t N = A.num_rows;

    residual.resize(N);
    y.resize(N);
    h.resize(N);
}

template <typename ValueType, typename MemorySpace>
template<typename MemorySpace2>
polynomial<ValueType,MemorySpace>
::polynomial(const polynomial<ValueType,MemorySpace2>& A) : default_coefficients(A.default_coefficients), residual(A.residual), h(A.h), y(A.y)
{
}

template <typename ValueType, typename MemorySpace>
template<typename MatrixType>
polynomial<ValueType,MemorySpace>
::polynomial(const cusp::precond::aggregation::sa_level<MatrixType>& sa_level)
{
    CUSP_PROFILE_SCOPED();

    size_t N = sa_level.A_.num_rows;

    ValueType rho = cusp::detail::ritz_spectral_radius_symmetric(sa_level.A_, 8);
    detail::chebyshev_polynomial_coefficients(rho, default_coefficients);
    default_coefficients.resize( default_coefficients.size() - 1 );

    for( size_t index = 0; index < default_coefficients.size(); index++ )
        default_coefficients[index] *= -1;

    residual.resize(N);
    y.resize(N);
    h.resize(N);
}

// linear_operator
template <typename ValueType, typename MemorySpace>
template<typename MatrixType, typename VectorType1, typename VectorType2>
void polynomial<ValueType,MemorySpace>
::operator()(const MatrixType& A, const VectorType1& b, VectorType2& x) const
{
    CUSP_PROFILE_SCOPED();

    polynomial<ValueType,MemorySpace>::operator()(A,b,x,default_coefficients);
}

template <typename ValueType, typename MemorySpace>
template<typename MatrixType, typename VectorType1, typename VectorType2>
void polynomial<ValueType,MemorySpace>
::presmooth(const MatrixType& A, const VectorType1& b, VectorType2& x)
{
    CUSP_PROFILE_SCOPED();

    // Ignore the initial x and use b as the residual
    ValueType scale_factor = default_coefficients[0];
    cusp::blas::axpby(b, x, x, scale_factor, ValueType(0));

    for( size_t i = 1; i<default_coefficients.size(); i++ )
    {
        scale_factor = default_coefficients[i];

        cusp::multiply(A, x, y);
        cusp::blas::axpby(y, b, x, ValueType(1.0), scale_factor);
    }
}

template <typename ValueType, typename MemorySpace>
template<typename MatrixType, typename VectorType1, typename VectorType2>
void polynomial<ValueType,MemorySpace>
::postsmooth(const MatrixType& A, const VectorType1& b, VectorType2& x)
{
    CUSP_PROFILE_SCOPED();

    // compute residual <- b - A*x
    cusp::multiply(A, x, residual);
    cusp::blas::axpby(b, residual, residual, ValueType(1), ValueType(-1));

    ValueType scale_factor = default_coefficients[0];
    cusp::blas::axpby(residual, h, h, scale_factor, ValueType(0));

    for( size_t i = 1; i<default_coefficients.size(); i++ )
    {
        scale_factor = default_coefficients[i];

        cusp::multiply(A, h, y);
        cusp::blas::axpby(y, residual, h, ValueType(1.0), scale_factor);
    }

    cusp::blas::axpy(h, x, ValueType(1.0));
}

// override default coefficients
template <typename ValueType, typename MemorySpace>
template<typename MatrixType, typename VectorType1, typename VectorType2, typename VectorType3>
void polynomial<ValueType,MemorySpace>
::operator()(const MatrixType& A, const VectorType1& b, VectorType2& x, VectorType3& coefficients)
{
    if( cusp::blas::nrm2(x) == 0.0 )
    {
        residual = b;
    }
    else
    {
        // compute residual <- b - A*x
        cusp::multiply(A, x, residual);
        cusp::blas::axpby(b, residual, residual, ValueType(1), ValueType(-1));
    }

    ValueType scale_factor = coefficients[0];
    cusp::blas::axpby(residual, h, h, scale_factor, ValueType(0));

    for( size_t i = 1; i<coefficients.size(); i++ )
    {
        scale_factor = coefficients[i];

        cusp::multiply(A, h, y);
        cusp::blas::axpby(y, residual, h, ValueType(1.0), scale_factor);
    }

    cusp::blas::axpy(h, x, ValueType(1.0));
}

} // end namespace relaxation
} // end namespace cusp