<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/>
<title>ublas: operation.hpp Source File</title>
<link href="tabs.css" rel="stylesheet" type="text/css"/>
<link href="doxygen.css" rel="stylesheet" type="text/css"/>
</head>
<body>
<!-- Generated by Doxygen 1.6.1 -->
<div class="navigation" id="top">
<div class="tabs">
<ul>
<li><a href="index.html"><span>Main Page</span></a></li>
<li><a href="pages.html"><span>Related Pages</span></a></li>
<li><a href="annotated.html"><span>Classes</span></a></li>
<li class="current"><a href="files.html"><span>Files</span></a></li>
</ul>
</div>
<div class="tabs">
<ul>
<li><a href="files.html"><span>File List</span></a></li>
</ul>
</div>
<h1>operation.hpp</h1><div class="fragment"><pre class="fragment"><a name="l00001"></a>00001 <span class="comment">//</span>
<a name="l00002"></a>00002 <span class="comment">// Copyright (c) 2000-2002</span>
<a name="l00003"></a>00003 <span class="comment">// Joerg Walter, Mathias Koch</span>
<a name="l00004"></a>00004 <span class="comment">//</span>
<a name="l00005"></a>00005 <span class="comment">// Distributed under the Boost Software License, Version 1.0. (See</span>
<a name="l00006"></a>00006 <span class="comment">// accompanying file LICENSE_1_0.txt or copy at</span>
<a name="l00007"></a>00007 <span class="comment">// http://www.boost.org/LICENSE_1_0.txt)</span>
<a name="l00008"></a>00008 <span class="comment">//</span>
<a name="l00009"></a>00009 <span class="comment">// The authors gratefully acknowledge the support of</span>
<a name="l00010"></a>00010 <span class="comment">// GeNeSys mbH & Co. KG in producing this work.</span>
<a name="l00011"></a>00011 <span class="comment">//</span>
<a name="l00012"></a>00012
<a name="l00013"></a>00013 <span class="preprocessor">#ifndef _BOOST_UBLAS_OPERATION_</span>
<a name="l00014"></a>00014 <span class="preprocessor"></span><span class="preprocessor">#define _BOOST_UBLAS_OPERATION_</span>
<a name="l00015"></a>00015 <span class="preprocessor"></span>
<a name="l00016"></a>00016 <span class="preprocessor">#include <boost/numeric/ublas/matrix_proxy.hpp></span>
<a name="l00017"></a>00017
<a name="l00022"></a>00022 <span class="comment">// axpy-based products</span>
<a name="l00023"></a>00023 <span class="comment">// Alexei Novakov had a lot of ideas to improve these. Thanks.</span>
<a name="l00024"></a>00024 <span class="comment">// Hendrik Kueck proposed some new kernel. Thanks again.</span>
<a name="l00025"></a>00025
<a name="l00026"></a>00026 <span class="keyword">namespace </span>boost { <span class="keyword">namespace </span>numeric { <span class="keyword">namespace </span>ublas {
<a name="l00027"></a>00027
<a name="l00028"></a>00028 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> T1, <span class="keyword">class</span> L1, <span class="keyword">class</span> IA1, <span class="keyword">class</span> TA1, <span class="keyword">class</span> E2>
<a name="l00029"></a>00029 BOOST_UBLAS_INLINE
<a name="l00030"></a>00030 V &
<a name="l00031"></a>00031 axpy_prod (<span class="keyword">const</span> compressed_matrix<T1, L1, 0, IA1, TA1> &e1,
<a name="l00032"></a>00032 <span class="keyword">const</span> vector_expression<E2> &e2,
<a name="l00033"></a>00033 V &v, row_major_tag) {
<a name="l00034"></a>00034 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::size_type size_type;
<a name="l00035"></a>00035 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::value_type value_type;
<a name="l00036"></a>00036
<a name="l00037"></a>00037 <span class="keywordflow">for</span> (size_type i = 0; i < e1.filled1 () -1; ++ i) {
<a name="l00038"></a>00038 size_type begin = e1.index1_data () [i];
<a name="l00039"></a>00039 size_type end = e1.index1_data () [i + 1];
<a name="l00040"></a>00040 value_type t (v (i));
<a name="l00041"></a>00041 <span class="keywordflow">for</span> (size_type j = begin; j < end; ++ j)
<a name="l00042"></a>00042 t += e1.value_data () [j] * e2 () (e1.index2_data () [j]);
<a name="l00043"></a>00043 v (i) = t;
<a name="l00044"></a>00044 }
<a name="l00045"></a>00045 <span class="keywordflow">return</span> v;
<a name="l00046"></a>00046 }
<a name="l00047"></a>00047
<a name="l00048"></a>00048 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> T1, <span class="keyword">class</span> L1, <span class="keyword">class</span> IA1, <span class="keyword">class</span> TA1, <span class="keyword">class</span> E2>
<a name="l00049"></a>00049 BOOST_UBLAS_INLINE
<a name="l00050"></a>00050 V &
<a name="l00051"></a>00051 axpy_prod (<span class="keyword">const</span> compressed_matrix<T1, L1, 0, IA1, TA1> &e1,
<a name="l00052"></a>00052 <span class="keyword">const</span> vector_expression<E2> &e2,
<a name="l00053"></a>00053 V &v, column_major_tag) {
<a name="l00054"></a>00054 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::size_type size_type;
<a name="l00055"></a>00055
<a name="l00056"></a>00056 <span class="keywordflow">for</span> (size_type j = 0; j < e1.filled1 () -1; ++ j) {
<a name="l00057"></a>00057 size_type begin = e1.index1_data () [j];
<a name="l00058"></a>00058 size_type end = e1.index1_data () [j + 1];
<a name="l00059"></a>00059 <span class="keywordflow">for</span> (size_type i = begin; i < end; ++ i)
<a name="l00060"></a>00060 v (e1.index2_data () [i]) += e1.value_data () [i] * e2 () (j);
<a name="l00061"></a>00061 }
<a name="l00062"></a>00062 <span class="keywordflow">return</span> v;
<a name="l00063"></a>00063 }
<a name="l00064"></a>00064
<a name="l00065"></a>00065 <span class="comment">// Dispatcher</span>
<a name="l00066"></a>00066 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> T1, <span class="keyword">class</span> L1, <span class="keyword">class</span> IA1, <span class="keyword">class</span> TA1, <span class="keyword">class</span> E2>
<a name="l00067"></a>00067 BOOST_UBLAS_INLINE
<a name="l00068"></a>00068 V &
<a name="l00069"></a>00069 axpy_prod (<span class="keyword">const</span> compressed_matrix<T1, L1, 0, IA1, TA1> &e1,
<a name="l00070"></a>00070 <span class="keyword">const</span> vector_expression<E2> &e2,
<a name="l00071"></a>00071 V &v, <span class="keywordtype">bool</span> init = <span class="keyword">true</span>) {
<a name="l00072"></a>00072 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::value_type value_type;
<a name="l00073"></a>00073 <span class="keyword">typedef</span> <span class="keyword">typename</span> L1::orientation_category orientation_category;
<a name="l00074"></a>00074
<a name="l00075"></a>00075 <span class="keywordflow">if</span> (init)
<a name="l00076"></a>00076 v.assign (zero_vector<value_type> (e1.size1 ()));
<a name="l00077"></a>00077 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00078"></a>00078 <span class="preprocessor"></span> vector<value_type> cv (v);
<a name="l00079"></a>00079 <span class="keyword">typedef</span> <span class="keyword">typename</span> type_traits<value_type>::real_type real_type;
<a name="l00080"></a>00080 real_type verrorbound (norm_1 (v) + norm_1 (e1) * norm_1 (e2));
<a name="l00081"></a>00081 indexing_vector_assign<scalar_plus_assign> (cv, prod (e1, e2));
<a name="l00082"></a>00082 <span class="preprocessor">#endif</span>
<a name="l00083"></a>00083 <span class="preprocessor"></span> axpy_prod (e1, e2, v, orientation_category ());
<a name="l00084"></a>00084 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00085"></a>00085 <span class="preprocessor"></span> BOOST_UBLAS_CHECK (norm_1 (v - cv) <= 2 * std::numeric_limits<real_type>::epsilon () * verrorbound, internal_logic ());
<a name="l00086"></a>00086 <span class="preprocessor">#endif</span>
<a name="l00087"></a>00087 <span class="preprocessor"></span> <span class="keywordflow">return</span> v;
<a name="l00088"></a>00088 }
<a name="l00089"></a>00089 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> T1, <span class="keyword">class</span> L1, <span class="keyword">class</span> IA1, <span class="keyword">class</span> TA1, <span class="keyword">class</span> E2>
<a name="l00090"></a>00090 BOOST_UBLAS_INLINE
<a name="l00091"></a>00091 V
<a name="l00092"></a>00092 axpy_prod (<span class="keyword">const</span> compressed_matrix<T1, L1, 0, IA1, TA1> &e1,
<a name="l00093"></a>00093 <span class="keyword">const</span> vector_expression<E2> &e2) {
<a name="l00094"></a>00094 <span class="keyword">typedef</span> V vector_type;
<a name="l00095"></a>00095
<a name="l00096"></a>00096 vector_type v (e1.size1 ());
<a name="l00097"></a>00097 <span class="keywordflow">return</span> axpy_prod (e1, e2, v, <span class="keyword">true</span>);
<a name="l00098"></a>00098 }
<a name="l00099"></a>00099
<a name="l00100"></a>00100 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> T1, <span class="keyword">class</span> L1, <span class="keyword">class</span> IA1, <span class="keyword">class</span> TA1, <span class="keyword">class</span> E2>
<a name="l00101"></a>00101 BOOST_UBLAS_INLINE
<a name="l00102"></a>00102 V &
<a name="l00103"></a>00103 axpy_prod (<span class="keyword">const</span> coordinate_matrix<T1, L1, 0, IA1, TA1> &e1,
<a name="l00104"></a>00104 <span class="keyword">const</span> vector_expression<E2> &e2,
<a name="l00105"></a>00105 V &v, <span class="keywordtype">bool</span> init = <span class="keyword">true</span>) {
<a name="l00106"></a>00106 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::size_type size_type;
<a name="l00107"></a>00107 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::value_type value_type;
<a name="l00108"></a>00108 <span class="keyword">typedef</span> L1 layout_type;
<a name="l00109"></a>00109
<a name="l00110"></a>00110 size_type size1 = e1.size1();
<a name="l00111"></a>00111 size_type size2 = e1.size2();
<a name="l00112"></a>00112
<a name="l00113"></a>00113 <span class="keywordflow">if</span> (init) {
<a name="l00114"></a>00114 noalias(v) = zero_vector<value_type>(size1);
<a name="l00115"></a>00115 }
<a name="l00116"></a>00116
<a name="l00117"></a>00117 <span class="keywordflow">for</span> (size_type i = 0; i < e1.nnz(); ++i) {
<a name="l00118"></a>00118 size_type row_index = layout_type::index_M( e1.index1_data () [i], e1.index2_data () [i] );
<a name="l00119"></a>00119 size_type col_index = layout_type::index_m( e1.index1_data () [i], e1.index2_data () [i] );
<a name="l00120"></a>00120 v( row_index ) += e1.value_data () [i] * e2 () (col_index);
<a name="l00121"></a>00121 }
<a name="l00122"></a>00122 <span class="keywordflow">return</span> v;
<a name="l00123"></a>00123 }
<a name="l00124"></a>00124
<a name="l00125"></a>00125 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00126"></a>00126 BOOST_UBLAS_INLINE
<a name="l00127"></a>00127 V &
<a name="l00128"></a>00128 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00129"></a>00129 <span class="keyword">const</span> vector_expression<E2> &e2,
<a name="l00130"></a>00130 V &v, packed_random_access_iterator_tag, row_major_tag) {
<a name="l00131"></a>00131 <span class="keyword">typedef</span> <span class="keyword">const</span> E1 expression1_type;
<a name="l00132"></a>00132 <span class="keyword">typedef</span> <span class="keyword">const</span> E2 expression2_type;
<a name="l00133"></a>00133 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::size_type size_type;
<a name="l00134"></a>00134
<a name="l00135"></a>00135 <span class="keyword">typename</span> expression1_type::const_iterator1 it1 (e1 ().begin1 ());
<a name="l00136"></a>00136 <span class="keyword">typename</span> expression1_type::const_iterator1 it1_end (e1 ().end1 ());
<a name="l00137"></a>00137 <span class="keywordflow">while</span> (it1 != it1_end) {
<a name="l00138"></a>00138 size_type index1 (it1.index1 ());
<a name="l00139"></a>00139 <span class="preprocessor">#ifndef BOOST_UBLAS_NO_NESTED_CLASS_RELATION</span>
<a name="l00140"></a>00140 <span class="preprocessor"></span> <span class="keyword">typename</span> expression1_type::const_iterator2 it2 (it1.begin ());
<a name="l00141"></a>00141 <span class="keyword">typename</span> expression1_type::const_iterator2 it2_end (it1.end ());
<a name="l00142"></a>00142 <span class="preprocessor">#else</span>
<a name="l00143"></a>00143 <span class="preprocessor"></span> <span class="keyword">typename</span> expression1_type::const_iterator2 it2 (boost::numeric::ublas::begin (it1, iterator1_tag ()));
<a name="l00144"></a>00144 <span class="keyword">typename</span> expression1_type::const_iterator2 it2_end (boost::numeric::ublas::end (it1, iterator1_tag ()));
<a name="l00145"></a>00145 <span class="preprocessor">#endif</span>
<a name="l00146"></a>00146 <span class="preprocessor"></span> <span class="keywordflow">while</span> (it2 != it2_end) {
<a name="l00147"></a>00147 v (index1) += *it2 * e2 () (it2.index2 ());
<a name="l00148"></a>00148 ++ it2;
<a name="l00149"></a>00149 }
<a name="l00150"></a>00150 ++ it1;
<a name="l00151"></a>00151 }
<a name="l00152"></a>00152 <span class="keywordflow">return</span> v;
<a name="l00153"></a>00153 }
<a name="l00154"></a>00154
<a name="l00155"></a>00155 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00156"></a>00156 BOOST_UBLAS_INLINE
<a name="l00157"></a>00157 V &
<a name="l00158"></a>00158 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00159"></a>00159 <span class="keyword">const</span> vector_expression<E2> &e2,
<a name="l00160"></a>00160 V &v, packed_random_access_iterator_tag, column_major_tag) {
<a name="l00161"></a>00161 <span class="keyword">typedef</span> <span class="keyword">const</span> E1 expression1_type;
<a name="l00162"></a>00162 <span class="keyword">typedef</span> <span class="keyword">const</span> E2 expression2_type;
<a name="l00163"></a>00163 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::size_type size_type;
<a name="l00164"></a>00164
<a name="l00165"></a>00165 <span class="keyword">typename</span> expression1_type::const_iterator2 it2 (e1 ().begin2 ());
<a name="l00166"></a>00166 <span class="keyword">typename</span> expression1_type::const_iterator2 it2_end (e1 ().end2 ());
<a name="l00167"></a>00167 <span class="keywordflow">while</span> (it2 != it2_end) {
<a name="l00168"></a>00168 size_type index2 (it2.index2 ());
<a name="l00169"></a>00169 <span class="preprocessor">#ifndef BOOST_UBLAS_NO_NESTED_CLASS_RELATION</span>
<a name="l00170"></a>00170 <span class="preprocessor"></span> <span class="keyword">typename</span> expression1_type::const_iterator1 it1 (it2.begin ());
<a name="l00171"></a>00171 <span class="keyword">typename</span> expression1_type::const_iterator1 it1_end (it2.end ());
<a name="l00172"></a>00172 <span class="preprocessor">#else</span>
<a name="l00173"></a>00173 <span class="preprocessor"></span> <span class="keyword">typename</span> expression1_type::const_iterator1 it1 (boost::numeric::ublas::begin (it2, iterator2_tag ()));
<a name="l00174"></a>00174 <span class="keyword">typename</span> expression1_type::const_iterator1 it1_end (boost::numeric::ublas::end (it2, iterator2_tag ()));
<a name="l00175"></a>00175 <span class="preprocessor">#endif</span>
<a name="l00176"></a>00176 <span class="preprocessor"></span> <span class="keywordflow">while</span> (it1 != it1_end) {
<a name="l00177"></a>00177 v (it1.index1 ()) += *it1 * e2 () (index2);
<a name="l00178"></a>00178 ++ it1;
<a name="l00179"></a>00179 }
<a name="l00180"></a>00180 ++ it2;
<a name="l00181"></a>00181 }
<a name="l00182"></a>00182 <span class="keywordflow">return</span> v;
<a name="l00183"></a>00183 }
<a name="l00184"></a>00184
<a name="l00185"></a>00185 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00186"></a>00186 BOOST_UBLAS_INLINE
<a name="l00187"></a>00187 V &
<a name="l00188"></a>00188 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00189"></a>00189 <span class="keyword">const</span> vector_expression<E2> &e2,
<a name="l00190"></a>00190 V &v, sparse_bidirectional_iterator_tag) {
<a name="l00191"></a>00191 <span class="keyword">typedef</span> <span class="keyword">const</span> E1 expression1_type;
<a name="l00192"></a>00192 <span class="keyword">typedef</span> <span class="keyword">const</span> E2 expression2_type;
<a name="l00193"></a>00193 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::size_type size_type;
<a name="l00194"></a>00194
<a name="l00195"></a>00195 <span class="keyword">typename</span> expression2_type::const_iterator it (e2 ().begin ());
<a name="l00196"></a>00196 <span class="keyword">typename</span> expression2_type::const_iterator it_end (e2 ().end ());
<a name="l00197"></a>00197 <span class="keywordflow">while</span> (it != it_end) {
<a name="l00198"></a>00198 v.plus_assign (column (e1 (), it.index ()) * *it);
<a name="l00199"></a>00199 ++ it;
<a name="l00200"></a>00200 }
<a name="l00201"></a>00201 <span class="keywordflow">return</span> v;
<a name="l00202"></a>00202 }
<a name="l00203"></a>00203
<a name="l00204"></a>00204 <span class="comment">// Dispatcher</span>
<a name="l00205"></a>00205 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00206"></a>00206 BOOST_UBLAS_INLINE
<a name="l00207"></a>00207 V &
<a name="l00208"></a>00208 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00209"></a>00209 <span class="keyword">const</span> vector_expression<E2> &e2,
<a name="l00210"></a>00210 V &v, packed_random_access_iterator_tag) {
<a name="l00211"></a>00211 <span class="keyword">typedef</span> <span class="keyword">typename</span> E1::orientation_category orientation_category;
<a name="l00212"></a>00212 <span class="keywordflow">return</span> axpy_prod (e1, e2, v, packed_random_access_iterator_tag (), orientation_category ());
<a name="l00213"></a>00213 }
<a name="l00214"></a>00214
<a name="l00215"></a>00215
<a name="l00241"></a>00241 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00242"></a>00242 BOOST_UBLAS_INLINE
<a name="l00243"></a>00243 V &
<a name="l00244"></a>00244 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00245"></a>00245 <span class="keyword">const</span> vector_expression<E2> &e2,
<a name="l00246"></a>00246 V &v, <span class="keywordtype">bool</span> init = <span class="keyword">true</span>) {
<a name="l00247"></a>00247 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::value_type value_type;
<a name="l00248"></a>00248 <span class="keyword">typedef</span> <span class="keyword">typename</span> E2::const_iterator::iterator_category iterator_category;
<a name="l00249"></a>00249
<a name="l00250"></a>00250 <span class="keywordflow">if</span> (init)
<a name="l00251"></a>00251 v.assign (zero_vector<value_type> (e1 ().size1 ()));
<a name="l00252"></a>00252 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00253"></a>00253 <span class="preprocessor"></span> vector<value_type> cv (v);
<a name="l00254"></a>00254 <span class="keyword">typedef</span> <span class="keyword">typename</span> type_traits<value_type>::real_type real_type;
<a name="l00255"></a>00255 real_type verrorbound (norm_1 (v) + norm_1 (e1) * norm_1 (e2));
<a name="l00256"></a>00256 indexing_vector_assign<scalar_plus_assign> (cv, prod (e1, e2));
<a name="l00257"></a>00257 <span class="preprocessor">#endif</span>
<a name="l00258"></a>00258 <span class="preprocessor"></span> axpy_prod (e1, e2, v, iterator_category ());
<a name="l00259"></a>00259 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00260"></a>00260 <span class="preprocessor"></span> BOOST_UBLAS_CHECK (norm_1 (v - cv) <= 2 * std::numeric_limits<real_type>::epsilon () * verrorbound, internal_logic ());
<a name="l00261"></a>00261 <span class="preprocessor">#endif</span>
<a name="l00262"></a>00262 <span class="preprocessor"></span> <span class="keywordflow">return</span> v;
<a name="l00263"></a>00263 }
<a name="l00264"></a>00264 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00265"></a>00265 BOOST_UBLAS_INLINE
<a name="l00266"></a>00266 V
<a name="l00267"></a>00267 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00268"></a>00268 <span class="keyword">const</span> vector_expression<E2> &e2) {
<a name="l00269"></a>00269 <span class="keyword">typedef</span> V vector_type;
<a name="l00270"></a>00270
<a name="l00271"></a>00271 vector_type v (e1 ().size1 ());
<a name="l00272"></a>00272 <span class="keywordflow">return</span> axpy_prod (e1, e2, v, <span class="keyword">true</span>);
<a name="l00273"></a>00273 }
<a name="l00274"></a>00274
<a name="l00275"></a>00275 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> T2, <span class="keyword">class</span> IA2, <span class="keyword">class</span> TA2>
<a name="l00276"></a>00276 BOOST_UBLAS_INLINE
<a name="l00277"></a>00277 V &
<a name="l00278"></a>00278 axpy_prod (<span class="keyword">const</span> vector_expression<E1> &e1,
<a name="l00279"></a>00279 <span class="keyword">const</span> compressed_matrix<T2, column_major, 0, IA2, TA2> &e2,
<a name="l00280"></a>00280 V &v, column_major_tag) {
<a name="l00281"></a>00281 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::size_type size_type;
<a name="l00282"></a>00282 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::value_type value_type;
<a name="l00283"></a>00283
<a name="l00284"></a>00284 <span class="keywordflow">for</span> (size_type j = 0; j < e2.filled1 () -1; ++ j) {
<a name="l00285"></a>00285 size_type begin = e2.index1_data () [j];
<a name="l00286"></a>00286 size_type end = e2.index1_data () [j + 1];
<a name="l00287"></a>00287 value_type t (v (j));
<a name="l00288"></a>00288 <span class="keywordflow">for</span> (size_type i = begin; i < end; ++ i)
<a name="l00289"></a>00289 t += e2.value_data () [i] * e1 () (e2.index2_data () [i]);
<a name="l00290"></a>00290 v (j) = t;
<a name="l00291"></a>00291 }
<a name="l00292"></a>00292 <span class="keywordflow">return</span> v;
<a name="l00293"></a>00293 }
<a name="l00294"></a>00294
<a name="l00295"></a>00295 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> T2, <span class="keyword">class</span> IA2, <span class="keyword">class</span> TA2>
<a name="l00296"></a>00296 BOOST_UBLAS_INLINE
<a name="l00297"></a>00297 V &
<a name="l00298"></a>00298 axpy_prod (<span class="keyword">const</span> vector_expression<E1> &e1,
<a name="l00299"></a>00299 <span class="keyword">const</span> compressed_matrix<T2, row_major, 0, IA2, TA2> &e2,
<a name="l00300"></a>00300 V &v, row_major_tag) {
<a name="l00301"></a>00301 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::size_type size_type;
<a name="l00302"></a>00302
<a name="l00303"></a>00303 <span class="keywordflow">for</span> (size_type i = 0; i < e2.filled1 () -1; ++ i) {
<a name="l00304"></a>00304 size_type begin = e2.index1_data () [i];
<a name="l00305"></a>00305 size_type end = e2.index1_data () [i + 1];
<a name="l00306"></a>00306 <span class="keywordflow">for</span> (size_type j = begin; j < end; ++ j)
<a name="l00307"></a>00307 v (e2.index2_data () [j]) += e2.value_data () [j] * e1 () (i);
<a name="l00308"></a>00308 }
<a name="l00309"></a>00309 <span class="keywordflow">return</span> v;
<a name="l00310"></a>00310 }
<a name="l00311"></a>00311
<a name="l00312"></a>00312 <span class="comment">// Dispatcher</span>
<a name="l00313"></a>00313 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> T2, <span class="keyword">class</span> L2, <span class="keyword">class</span> IA2, <span class="keyword">class</span> TA2>
<a name="l00314"></a>00314 BOOST_UBLAS_INLINE
<a name="l00315"></a>00315 V &
<a name="l00316"></a>00316 axpy_prod (<span class="keyword">const</span> vector_expression<E1> &e1,
<a name="l00317"></a>00317 <span class="keyword">const</span> compressed_matrix<T2, L2, 0, IA2, TA2> &e2,
<a name="l00318"></a>00318 V &v, <span class="keywordtype">bool</span> init = <span class="keyword">true</span>) {
<a name="l00319"></a>00319 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::value_type value_type;
<a name="l00320"></a>00320 <span class="keyword">typedef</span> <span class="keyword">typename</span> L2::orientation_category orientation_category;
<a name="l00321"></a>00321
<a name="l00322"></a>00322 <span class="keywordflow">if</span> (init)
<a name="l00323"></a>00323 v.assign (zero_vector<value_type> (e2.size2 ()));
<a name="l00324"></a>00324 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00325"></a>00325 <span class="preprocessor"></span> vector<value_type> cv (v);
<a name="l00326"></a>00326 <span class="keyword">typedef</span> <span class="keyword">typename</span> type_traits<value_type>::real_type real_type;
<a name="l00327"></a>00327 real_type verrorbound (norm_1 (v) + norm_1 (e1) * norm_1 (e2));
<a name="l00328"></a>00328 indexing_vector_assign<scalar_plus_assign> (cv, prod (e1, e2));
<a name="l00329"></a>00329 <span class="preprocessor">#endif</span>
<a name="l00330"></a>00330 <span class="preprocessor"></span> axpy_prod (e1, e2, v, orientation_category ());
<a name="l00331"></a>00331 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00332"></a>00332 <span class="preprocessor"></span> BOOST_UBLAS_CHECK (norm_1 (v - cv) <= 2 * std::numeric_limits<real_type>::epsilon () * verrorbound, internal_logic ());
<a name="l00333"></a>00333 <span class="preprocessor">#endif</span>
<a name="l00334"></a>00334 <span class="preprocessor"></span> <span class="keywordflow">return</span> v;
<a name="l00335"></a>00335 }
<a name="l00336"></a>00336 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> T2, <span class="keyword">class</span> L2, <span class="keyword">class</span> IA2, <span class="keyword">class</span> TA2>
<a name="l00337"></a>00337 BOOST_UBLAS_INLINE
<a name="l00338"></a>00338 V
<a name="l00339"></a>00339 axpy_prod (<span class="keyword">const</span> vector_expression<E1> &e1,
<a name="l00340"></a>00340 <span class="keyword">const</span> compressed_matrix<T2, L2, 0, IA2, TA2> &e2) {
<a name="l00341"></a>00341 <span class="keyword">typedef</span> V vector_type;
<a name="l00342"></a>00342
<a name="l00343"></a>00343 vector_type v (e2.size2 ());
<a name="l00344"></a>00344 <span class="keywordflow">return</span> axpy_prod (e1, e2, v, <span class="keyword">true</span>);
<a name="l00345"></a>00345 }
<a name="l00346"></a>00346
<a name="l00347"></a>00347 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00348"></a>00348 BOOST_UBLAS_INLINE
<a name="l00349"></a>00349 V &
<a name="l00350"></a>00350 axpy_prod (<span class="keyword">const</span> vector_expression<E1> &e1,
<a name="l00351"></a>00351 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00352"></a>00352 V &v, packed_random_access_iterator_tag, column_major_tag) {
<a name="l00353"></a>00353 <span class="keyword">typedef</span> <span class="keyword">const</span> E1 expression1_type;
<a name="l00354"></a>00354 <span class="keyword">typedef</span> <span class="keyword">const</span> E2 expression2_type;
<a name="l00355"></a>00355 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::size_type size_type;
<a name="l00356"></a>00356
<a name="l00357"></a>00357 <span class="keyword">typename</span> expression2_type::const_iterator2 it2 (e2 ().begin2 ());
<a name="l00358"></a>00358 <span class="keyword">typename</span> expression2_type::const_iterator2 it2_end (e2 ().end2 ());
<a name="l00359"></a>00359 <span class="keywordflow">while</span> (it2 != it2_end) {
<a name="l00360"></a>00360 size_type index2 (it2.index2 ());
<a name="l00361"></a>00361 <span class="preprocessor">#ifndef BOOST_UBLAS_NO_NESTED_CLASS_RELATION</span>
<a name="l00362"></a>00362 <span class="preprocessor"></span> <span class="keyword">typename</span> expression2_type::const_iterator1 it1 (it2.begin ());
<a name="l00363"></a>00363 <span class="keyword">typename</span> expression2_type::const_iterator1 it1_end (it2.end ());
<a name="l00364"></a>00364 <span class="preprocessor">#else</span>
<a name="l00365"></a>00365 <span class="preprocessor"></span> <span class="keyword">typename</span> expression2_type::const_iterator1 it1 (boost::numeric::ublas::begin (it2, iterator2_tag ()));
<a name="l00366"></a>00366 <span class="keyword">typename</span> expression2_type::const_iterator1 it1_end (boost::numeric::ublas::end (it2, iterator2_tag ()));
<a name="l00367"></a>00367 <span class="preprocessor">#endif</span>
<a name="l00368"></a>00368 <span class="preprocessor"></span> <span class="keywordflow">while</span> (it1 != it1_end) {
<a name="l00369"></a>00369 v (index2) += *it1 * e1 () (it1.index1 ());
<a name="l00370"></a>00370 ++ it1;
<a name="l00371"></a>00371 }
<a name="l00372"></a>00372 ++ it2;
<a name="l00373"></a>00373 }
<a name="l00374"></a>00374 <span class="keywordflow">return</span> v;
<a name="l00375"></a>00375 }
<a name="l00376"></a>00376
<a name="l00377"></a>00377 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00378"></a>00378 BOOST_UBLAS_INLINE
<a name="l00379"></a>00379 V &
<a name="l00380"></a>00380 axpy_prod (<span class="keyword">const</span> vector_expression<E1> &e1,
<a name="l00381"></a>00381 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00382"></a>00382 V &v, packed_random_access_iterator_tag, row_major_tag) {
<a name="l00383"></a>00383 <span class="keyword">typedef</span> <span class="keyword">const</span> E1 expression1_type;
<a name="l00384"></a>00384 <span class="keyword">typedef</span> <span class="keyword">const</span> E2 expression2_type;
<a name="l00385"></a>00385 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::size_type size_type;
<a name="l00386"></a>00386
<a name="l00387"></a>00387 <span class="keyword">typename</span> expression2_type::const_iterator1 it1 (e2 ().begin1 ());
<a name="l00388"></a>00388 <span class="keyword">typename</span> expression2_type::const_iterator1 it1_end (e2 ().end1 ());
<a name="l00389"></a>00389 <span class="keywordflow">while</span> (it1 != it1_end) {
<a name="l00390"></a>00390 size_type index1 (it1.index1 ());
<a name="l00391"></a>00391 <span class="preprocessor">#ifndef BOOST_UBLAS_NO_NESTED_CLASS_RELATION</span>
<a name="l00392"></a>00392 <span class="preprocessor"></span> <span class="keyword">typename</span> expression2_type::const_iterator2 it2 (it1.begin ());
<a name="l00393"></a>00393 <span class="keyword">typename</span> expression2_type::const_iterator2 it2_end (it1.end ());
<a name="l00394"></a>00394 <span class="preprocessor">#else</span>
<a name="l00395"></a>00395 <span class="preprocessor"></span> <span class="keyword">typename</span> expression2_type::const_iterator2 it2 (boost::numeric::ublas::begin (it1, iterator1_tag ()));
<a name="l00396"></a>00396 <span class="keyword">typename</span> expression2_type::const_iterator2 it2_end (boost::numeric::ublas::end (it1, iterator1_tag ()));
<a name="l00397"></a>00397 <span class="preprocessor">#endif</span>
<a name="l00398"></a>00398 <span class="preprocessor"></span> <span class="keywordflow">while</span> (it2 != it2_end) {
<a name="l00399"></a>00399 v (it2.index2 ()) += *it2 * e1 () (index1);
<a name="l00400"></a>00400 ++ it2;
<a name="l00401"></a>00401 }
<a name="l00402"></a>00402 ++ it1;
<a name="l00403"></a>00403 }
<a name="l00404"></a>00404 <span class="keywordflow">return</span> v;
<a name="l00405"></a>00405 }
<a name="l00406"></a>00406
<a name="l00407"></a>00407 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00408"></a>00408 BOOST_UBLAS_INLINE
<a name="l00409"></a>00409 V &
<a name="l00410"></a>00410 axpy_prod (<span class="keyword">const</span> vector_expression<E1> &e1,
<a name="l00411"></a>00411 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00412"></a>00412 V &v, sparse_bidirectional_iterator_tag) {
<a name="l00413"></a>00413 <span class="keyword">typedef</span> <span class="keyword">const</span> E1 expression1_type;
<a name="l00414"></a>00414 <span class="keyword">typedef</span> <span class="keyword">const</span> E2 expression2_type;
<a name="l00415"></a>00415 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::size_type size_type;
<a name="l00416"></a>00416
<a name="l00417"></a>00417 <span class="keyword">typename</span> expression1_type::const_iterator it (e1 ().begin ());
<a name="l00418"></a>00418 <span class="keyword">typename</span> expression1_type::const_iterator it_end (e1 ().end ());
<a name="l00419"></a>00419 <span class="keywordflow">while</span> (it != it_end) {
<a name="l00420"></a>00420 v.plus_assign (*it * row (e2 (), it.index ()));
<a name="l00421"></a>00421 ++ it;
<a name="l00422"></a>00422 }
<a name="l00423"></a>00423 <span class="keywordflow">return</span> v;
<a name="l00424"></a>00424 }
<a name="l00425"></a>00425
<a name="l00426"></a>00426 <span class="comment">// Dispatcher</span>
<a name="l00427"></a>00427 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00428"></a>00428 BOOST_UBLAS_INLINE
<a name="l00429"></a>00429 V &
<a name="l00430"></a>00430 axpy_prod (<span class="keyword">const</span> vector_expression<E1> &e1,
<a name="l00431"></a>00431 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00432"></a>00432 V &v, packed_random_access_iterator_tag) {
<a name="l00433"></a>00433 <span class="keyword">typedef</span> <span class="keyword">typename</span> E2::orientation_category orientation_category;
<a name="l00434"></a>00434 <span class="keywordflow">return</span> axpy_prod (e1, e2, v, packed_random_access_iterator_tag (), orientation_category ());
<a name="l00435"></a>00435 }
<a name="l00436"></a>00436
<a name="l00437"></a>00437
<a name="l00463"></a>00463 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00464"></a>00464 BOOST_UBLAS_INLINE
<a name="l00465"></a>00465 V &
<a name="l00466"></a>00466 axpy_prod (<span class="keyword">const</span> vector_expression<E1> &e1,
<a name="l00467"></a>00467 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00468"></a>00468 V &v, <span class="keywordtype">bool</span> init = <span class="keyword">true</span>) {
<a name="l00469"></a>00469 <span class="keyword">typedef</span> <span class="keyword">typename</span> V::value_type value_type;
<a name="l00470"></a>00470 <span class="keyword">typedef</span> <span class="keyword">typename</span> E1::const_iterator::iterator_category iterator_category;
<a name="l00471"></a>00471
<a name="l00472"></a>00472 <span class="keywordflow">if</span> (init)
<a name="l00473"></a>00473 v.assign (zero_vector<value_type> (e2 ().size2 ()));
<a name="l00474"></a>00474 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00475"></a>00475 <span class="preprocessor"></span> vector<value_type> cv (v);
<a name="l00476"></a>00476 <span class="keyword">typedef</span> <span class="keyword">typename</span> type_traits<value_type>::real_type real_type;
<a name="l00477"></a>00477 real_type verrorbound (norm_1 (v) + norm_1 (e1) * norm_1 (e2));
<a name="l00478"></a>00478 indexing_vector_assign<scalar_plus_assign> (cv, prod (e1, e2));
<a name="l00479"></a>00479 <span class="preprocessor">#endif</span>
<a name="l00480"></a>00480 <span class="preprocessor"></span> axpy_prod (e1, e2, v, iterator_category ());
<a name="l00481"></a>00481 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00482"></a>00482 <span class="preprocessor"></span> BOOST_UBLAS_CHECK (norm_1 (v - cv) <= 2 * std::numeric_limits<real_type>::epsilon () * verrorbound, internal_logic ());
<a name="l00483"></a>00483 <span class="preprocessor">#endif</span>
<a name="l00484"></a>00484 <span class="preprocessor"></span> <span class="keywordflow">return</span> v;
<a name="l00485"></a>00485 }
<a name="l00486"></a>00486 <span class="keyword">template</span><<span class="keyword">class</span> V, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00487"></a>00487 BOOST_UBLAS_INLINE
<a name="l00488"></a>00488 V
<a name="l00489"></a>00489 axpy_prod (<span class="keyword">const</span> vector_expression<E1> &e1,
<a name="l00490"></a>00490 <span class="keyword">const</span> matrix_expression<E2> &e2) {
<a name="l00491"></a>00491 <span class="keyword">typedef</span> V vector_type;
<a name="l00492"></a>00492
<a name="l00493"></a>00493 vector_type v (e2 ().size2 ());
<a name="l00494"></a>00494 <span class="keywordflow">return</span> axpy_prod (e1, e2, v, <span class="keyword">true</span>);
<a name="l00495"></a>00495 }
<a name="l00496"></a>00496
<a name="l00497"></a>00497 <span class="keyword">template</span><<span class="keyword">class</span> M, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2, <span class="keyword">class</span> TRI>
<a name="l00498"></a>00498 BOOST_UBLAS_INLINE
<a name="l00499"></a>00499 M &
<a name="l00500"></a>00500 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00501"></a>00501 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00502"></a>00502 M &m, TRI,
<a name="l00503"></a>00503 dense_proxy_tag, row_major_tag) {
<a name="l00504"></a>00504 <span class="keyword">typedef</span> M matrix_type;
<a name="l00505"></a>00505 <span class="keyword">typedef</span> <span class="keyword">const</span> E1 expression1_type;
<a name="l00506"></a>00506 <span class="keyword">typedef</span> <span class="keyword">const</span> E2 expression2_type;
<a name="l00507"></a>00507 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::size_type size_type;
<a name="l00508"></a>00508 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::value_type value_type;
<a name="l00509"></a>00509
<a name="l00510"></a>00510 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00511"></a>00511 <span class="preprocessor"></span> matrix<value_type, row_major> cm (m);
<a name="l00512"></a>00512 <span class="keyword">typedef</span> <span class="keyword">typename</span> type_traits<value_type>::real_type real_type;
<a name="l00513"></a>00513 real_type merrorbound (norm_1 (m) + norm_1 (e1) * norm_1 (e2));
<a name="l00514"></a>00514 indexing_matrix_assign<scalar_plus_assign> (cm, prod (e1, e2), row_major_tag ());
<a name="l00515"></a>00515 <span class="preprocessor">#endif</span>
<a name="l00516"></a>00516 <span class="preprocessor"></span> size_type size1 (e1 ().size1 ());
<a name="l00517"></a>00517 size_type size2 (e1 ().size2 ());
<a name="l00518"></a>00518 <span class="keywordflow">for</span> (size_type i = 0; i < size1; ++ i)
<a name="l00519"></a>00519 <span class="keywordflow">for</span> (size_type j = 0; j < size2; ++ j)
<a name="l00520"></a>00520 row (m, i).plus_assign (e1 () (i, j) * row (e2 (), j));
<a name="l00521"></a>00521 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00522"></a>00522 <span class="preprocessor"></span> BOOST_UBLAS_CHECK (norm_1 (m - cm) <= 2 * std::numeric_limits<real_type>::epsilon () * merrorbound, internal_logic ());
<a name="l00523"></a>00523 <span class="preprocessor">#endif</span>
<a name="l00524"></a>00524 <span class="preprocessor"></span> <span class="keywordflow">return</span> m;
<a name="l00525"></a>00525 }
<a name="l00526"></a>00526 <span class="keyword">template</span><<span class="keyword">class</span> M, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2, <span class="keyword">class</span> TRI>
<a name="l00527"></a>00527 BOOST_UBLAS_INLINE
<a name="l00528"></a>00528 M &
<a name="l00529"></a>00529 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00530"></a>00530 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00531"></a>00531 M &m, TRI,
<a name="l00532"></a>00532 sparse_proxy_tag, row_major_tag) {
<a name="l00533"></a>00533 <span class="keyword">typedef</span> M matrix_type;
<a name="l00534"></a>00534 <span class="keyword">typedef</span> TRI triangular_restriction;
<a name="l00535"></a>00535 <span class="keyword">typedef</span> <span class="keyword">const</span> E1 expression1_type;
<a name="l00536"></a>00536 <span class="keyword">typedef</span> <span class="keyword">const</span> E2 expression2_type;
<a name="l00537"></a>00537 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::size_type size_type;
<a name="l00538"></a>00538 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::value_type value_type;
<a name="l00539"></a>00539
<a name="l00540"></a>00540 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00541"></a>00541 <span class="preprocessor"></span> matrix<value_type, row_major> cm (m);
<a name="l00542"></a>00542 <span class="keyword">typedef</span> <span class="keyword">typename</span> type_traits<value_type>::real_type real_type;
<a name="l00543"></a>00543 real_type merrorbound (norm_1 (m) + norm_1 (e1) * norm_1 (e2));
<a name="l00544"></a>00544 indexing_matrix_assign<scalar_plus_assign> (cm, prod (e1, e2), row_major_tag ());
<a name="l00545"></a>00545 <span class="preprocessor">#endif</span>
<a name="l00546"></a>00546 <span class="preprocessor"></span> <span class="keyword">typename</span> expression1_type::const_iterator1 it1 (e1 ().begin1 ());
<a name="l00547"></a>00547 <span class="keyword">typename</span> expression1_type::const_iterator1 it1_end (e1 ().end1 ());
<a name="l00548"></a>00548 <span class="keywordflow">while</span> (it1 != it1_end) {
<a name="l00549"></a>00549 <span class="preprocessor">#ifndef BOOST_UBLAS_NO_NESTED_CLASS_RELATION</span>
<a name="l00550"></a>00550 <span class="preprocessor"></span> <span class="keyword">typename</span> expression1_type::const_iterator2 it2 (it1.begin ());
<a name="l00551"></a>00551 <span class="keyword">typename</span> expression1_type::const_iterator2 it2_end (it1.end ());
<a name="l00552"></a>00552 <span class="preprocessor">#else</span>
<a name="l00553"></a>00553 <span class="preprocessor"></span> <span class="keyword">typename</span> expression1_type::const_iterator2 it2 (boost::numeric::ublas::begin (it1, iterator1_tag ()));
<a name="l00554"></a>00554 <span class="keyword">typename</span> expression1_type::const_iterator2 it2_end (boost::numeric::ublas::end (it1, iterator1_tag ()));
<a name="l00555"></a>00555 <span class="preprocessor">#endif</span>
<a name="l00556"></a>00556 <span class="preprocessor"></span> <span class="keywordflow">while</span> (it2 != it2_end) {
<a name="l00557"></a>00557 <span class="comment">// row (m, it1.index1 ()).plus_assign (*it2 * row (e2 (), it2.index2 ()));</span>
<a name="l00558"></a>00558 matrix_row<expression2_type> mr (e2 (), it2.index2 ());
<a name="l00559"></a>00559 <span class="keyword">typename</span> matrix_row<expression2_type>::const_iterator itr (mr.begin ());
<a name="l00560"></a>00560 <span class="keyword">typename</span> matrix_row<expression2_type>::const_iterator itr_end (mr.end ());
<a name="l00561"></a>00561 <span class="keywordflow">while</span> (itr != itr_end) {
<a name="l00562"></a>00562 <span class="keywordflow">if</span> (triangular_restriction::other (it1.index1 (), itr.index ()))
<a name="l00563"></a>00563 m (it1.index1 (), itr.index ()) += *it2 * *itr;
<a name="l00564"></a>00564 ++ itr;
<a name="l00565"></a>00565 }
<a name="l00566"></a>00566 ++ it2;
<a name="l00567"></a>00567 }
<a name="l00568"></a>00568 ++ it1;
<a name="l00569"></a>00569 }
<a name="l00570"></a>00570 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00571"></a>00571 <span class="preprocessor"></span> BOOST_UBLAS_CHECK (norm_1 (m - cm) <= 2 * std::numeric_limits<real_type>::epsilon () * merrorbound, internal_logic ());
<a name="l00572"></a>00572 <span class="preprocessor">#endif</span>
<a name="l00573"></a>00573 <span class="preprocessor"></span> <span class="keywordflow">return</span> m;
<a name="l00574"></a>00574 }
<a name="l00575"></a>00575
<a name="l00576"></a>00576 <span class="keyword">template</span><<span class="keyword">class</span> M, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2, <span class="keyword">class</span> TRI>
<a name="l00577"></a>00577 BOOST_UBLAS_INLINE
<a name="l00578"></a>00578 M &
<a name="l00579"></a>00579 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00580"></a>00580 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00581"></a>00581 M &m, TRI,
<a name="l00582"></a>00582 dense_proxy_tag, column_major_tag) {
<a name="l00583"></a>00583 <span class="keyword">typedef</span> M matrix_type;
<a name="l00584"></a>00584 <span class="keyword">typedef</span> <span class="keyword">const</span> E1 expression1_type;
<a name="l00585"></a>00585 <span class="keyword">typedef</span> <span class="keyword">const</span> E2 expression2_type;
<a name="l00586"></a>00586 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::size_type size_type;
<a name="l00587"></a>00587 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::value_type value_type;
<a name="l00588"></a>00588
<a name="l00589"></a>00589 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00590"></a>00590 <span class="preprocessor"></span> matrix<value_type, column_major> cm (m);
<a name="l00591"></a>00591 <span class="keyword">typedef</span> <span class="keyword">typename</span> type_traits<value_type>::real_type real_type;
<a name="l00592"></a>00592 real_type merrorbound (norm_1 (m) + norm_1 (e1) * norm_1 (e2));
<a name="l00593"></a>00593 indexing_matrix_assign<scalar_plus_assign> (cm, prod (e1, e2), column_major_tag ());
<a name="l00594"></a>00594 <span class="preprocessor">#endif</span>
<a name="l00595"></a>00595 <span class="preprocessor"></span> size_type size1 (e2 ().size1 ());
<a name="l00596"></a>00596 size_type size2 (e2 ().size2 ());
<a name="l00597"></a>00597 <span class="keywordflow">for</span> (size_type j = 0; j < size2; ++ j)
<a name="l00598"></a>00598 <span class="keywordflow">for</span> (size_type i = 0; i < size1; ++ i)
<a name="l00599"></a>00599 column (m, j).plus_assign (e2 () (i, j) * column (e1 (), i));
<a name="l00600"></a>00600 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00601"></a>00601 <span class="preprocessor"></span> BOOST_UBLAS_CHECK (norm_1 (m - cm) <= 2 * std::numeric_limits<real_type>::epsilon () * merrorbound, internal_logic ());
<a name="l00602"></a>00602 <span class="preprocessor">#endif</span>
<a name="l00603"></a>00603 <span class="preprocessor"></span> <span class="keywordflow">return</span> m;
<a name="l00604"></a>00604 }
<a name="l00605"></a>00605 <span class="keyword">template</span><<span class="keyword">class</span> M, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2, <span class="keyword">class</span> TRI>
<a name="l00606"></a>00606 BOOST_UBLAS_INLINE
<a name="l00607"></a>00607 M &
<a name="l00608"></a>00608 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00609"></a>00609 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00610"></a>00610 M &m, TRI,
<a name="l00611"></a>00611 sparse_proxy_tag, column_major_tag) {
<a name="l00612"></a>00612 <span class="keyword">typedef</span> M matrix_type;
<a name="l00613"></a>00613 <span class="keyword">typedef</span> TRI triangular_restriction;
<a name="l00614"></a>00614 <span class="keyword">typedef</span> <span class="keyword">const</span> E1 expression1_type;
<a name="l00615"></a>00615 <span class="keyword">typedef</span> <span class="keyword">const</span> E2 expression2_type;
<a name="l00616"></a>00616 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::size_type size_type;
<a name="l00617"></a>00617 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::value_type value_type;
<a name="l00618"></a>00618
<a name="l00619"></a>00619 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00620"></a>00620 <span class="preprocessor"></span> matrix<value_type, column_major> cm (m);
<a name="l00621"></a>00621 <span class="keyword">typedef</span> <span class="keyword">typename</span> type_traits<value_type>::real_type real_type;
<a name="l00622"></a>00622 real_type merrorbound (norm_1 (m) + norm_1 (e1) * norm_1 (e2));
<a name="l00623"></a>00623 indexing_matrix_assign<scalar_plus_assign> (cm, prod (e1, e2), column_major_tag ());
<a name="l00624"></a>00624 <span class="preprocessor">#endif</span>
<a name="l00625"></a>00625 <span class="preprocessor"></span> <span class="keyword">typename</span> expression2_type::const_iterator2 it2 (e2 ().begin2 ());
<a name="l00626"></a>00626 <span class="keyword">typename</span> expression2_type::const_iterator2 it2_end (e2 ().end2 ());
<a name="l00627"></a>00627 <span class="keywordflow">while</span> (it2 != it2_end) {
<a name="l00628"></a>00628 <span class="preprocessor">#ifndef BOOST_UBLAS_NO_NESTED_CLASS_RELATION</span>
<a name="l00629"></a>00629 <span class="preprocessor"></span> <span class="keyword">typename</span> expression2_type::const_iterator1 it1 (it2.begin ());
<a name="l00630"></a>00630 <span class="keyword">typename</span> expression2_type::const_iterator1 it1_end (it2.end ());
<a name="l00631"></a>00631 <span class="preprocessor">#else</span>
<a name="l00632"></a>00632 <span class="preprocessor"></span> <span class="keyword">typename</span> expression2_type::const_iterator1 it1 (boost::numeric::ublas::begin (it2, iterator2_tag ()));
<a name="l00633"></a>00633 <span class="keyword">typename</span> expression2_type::const_iterator1 it1_end (boost::numeric::ublas::end (it2, iterator2_tag ()));
<a name="l00634"></a>00634 <span class="preprocessor">#endif</span>
<a name="l00635"></a>00635 <span class="preprocessor"></span> <span class="keywordflow">while</span> (it1 != it1_end) {
<a name="l00636"></a>00636 <span class="comment">// column (m, it2.index2 ()).plus_assign (*it1 * column (e1 (), it1.index1 ()));</span>
<a name="l00637"></a>00637 matrix_column<expression1_type> mc (e1 (), it1.index1 ());
<a name="l00638"></a>00638 <span class="keyword">typename</span> matrix_column<expression1_type>::const_iterator itc (mc.begin ());
<a name="l00639"></a>00639 <span class="keyword">typename</span> matrix_column<expression1_type>::const_iterator itc_end (mc.end ());
<a name="l00640"></a>00640 <span class="keywordflow">while</span> (itc != itc_end) {
<a name="l00641"></a>00641 <span class="keywordflow">if</span>(triangular_restriction::other (itc.index (), it2.index2 ()))
<a name="l00642"></a>00642 m (itc.index (), it2.index2 ()) += *it1 * *itc;
<a name="l00643"></a>00643 ++ itc;
<a name="l00644"></a>00644 }
<a name="l00645"></a>00645 ++ it1;
<a name="l00646"></a>00646 }
<a name="l00647"></a>00647 ++ it2;
<a name="l00648"></a>00648 }
<a name="l00649"></a>00649 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00650"></a>00650 <span class="preprocessor"></span> BOOST_UBLAS_CHECK (norm_1 (m - cm) <= 2 * std::numeric_limits<real_type>::epsilon () * merrorbound, internal_logic ());
<a name="l00651"></a>00651 <span class="preprocessor">#endif</span>
<a name="l00652"></a>00652 <span class="preprocessor"></span> <span class="keywordflow">return</span> m;
<a name="l00653"></a>00653 }
<a name="l00654"></a>00654
<a name="l00655"></a>00655 <span class="comment">// Dispatcher</span>
<a name="l00656"></a>00656 <span class="keyword">template</span><<span class="keyword">class</span> M, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2, <span class="keyword">class</span> TRI>
<a name="l00657"></a>00657 BOOST_UBLAS_INLINE
<a name="l00658"></a>00658 M &
<a name="l00659"></a>00659 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00660"></a>00660 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00661"></a>00661 M &m, TRI, <span class="keywordtype">bool</span> init = <span class="keyword">true</span>) {
<a name="l00662"></a>00662 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::value_type value_type;
<a name="l00663"></a>00663 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::storage_category storage_category;
<a name="l00664"></a>00664 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::orientation_category orientation_category;
<a name="l00665"></a>00665 <span class="keyword">typedef</span> TRI triangular_restriction;
<a name="l00666"></a>00666
<a name="l00667"></a>00667 <span class="keywordflow">if</span> (init)
<a name="l00668"></a>00668 m.assign (zero_matrix<value_type> (e1 ().size1 (), e2 ().size2 ()));
<a name="l00669"></a>00669 <span class="keywordflow">return</span> axpy_prod (e1, e2, m, triangular_restriction (), storage_category (), orientation_category ());
<a name="l00670"></a>00670 }
<a name="l00671"></a>00671 <span class="keyword">template</span><<span class="keyword">class</span> M, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2, <span class="keyword">class</span> TRI>
<a name="l00672"></a>00672 BOOST_UBLAS_INLINE
<a name="l00673"></a>00673 M
<a name="l00674"></a>00674 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00675"></a>00675 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00676"></a>00676 TRI) {
<a name="l00677"></a>00677 <span class="keyword">typedef</span> M matrix_type;
<a name="l00678"></a>00678 <span class="keyword">typedef</span> TRI triangular_restriction;
<a name="l00679"></a>00679
<a name="l00680"></a>00680 matrix_type m (e1 ().size1 (), e2 ().size2 ());
<a name="l00681"></a>00681 <span class="keywordflow">return</span> axpy_prod (e1, e2, m, triangular_restriction (), <span class="keyword">true</span>);
<a name="l00682"></a>00682 }
<a name="l00683"></a>00683
<a name="l00708"></a>00708 <span class="keyword">template</span><<span class="keyword">class</span> M, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00709"></a>00709 BOOST_UBLAS_INLINE
<a name="l00710"></a>00710 M &
<a name="l00711"></a>00711 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00712"></a>00712 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00713"></a>00713 M &m, <span class="keywordtype">bool</span> init = <span class="keyword">true</span>) {
<a name="l00714"></a>00714 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::value_type value_type;
<a name="l00715"></a>00715 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::storage_category storage_category;
<a name="l00716"></a>00716 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::orientation_category orientation_category;
<a name="l00717"></a>00717
<a name="l00718"></a>00718 <span class="keywordflow">if</span> (init)
<a name="l00719"></a>00719 m.assign (zero_matrix<value_type> (e1 ().size1 (), e2 ().size2 ()));
<a name="l00720"></a>00720 <span class="keywordflow">return</span> axpy_prod (e1, e2, m, full (), storage_category (), orientation_category ());
<a name="l00721"></a>00721 }
<a name="l00722"></a>00722 <span class="keyword">template</span><<span class="keyword">class</span> M, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00723"></a>00723 BOOST_UBLAS_INLINE
<a name="l00724"></a>00724 M
<a name="l00725"></a>00725 axpy_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00726"></a>00726 <span class="keyword">const</span> matrix_expression<E2> &e2) {
<a name="l00727"></a>00727 <span class="keyword">typedef</span> M matrix_type;
<a name="l00728"></a>00728
<a name="l00729"></a>00729 matrix_type m (e1 ().size1 (), e2 ().size2 ());
<a name="l00730"></a>00730 <span class="keywordflow">return</span> axpy_prod (e1, e2, m, full (), <span class="keyword">true</span>);
<a name="l00731"></a>00731 }
<a name="l00732"></a>00732
<a name="l00733"></a>00733
<a name="l00734"></a>00734 <span class="keyword">template</span><<span class="keyword">class</span> M, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00735"></a>00735 BOOST_UBLAS_INLINE
<a name="l00736"></a>00736 M &
<a name="l00737"></a>00737 opb_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00738"></a>00738 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00739"></a>00739 M &m,
<a name="l00740"></a>00740 dense_proxy_tag, row_major_tag) {
<a name="l00741"></a>00741 <span class="keyword">typedef</span> M matrix_type;
<a name="l00742"></a>00742 <span class="keyword">typedef</span> <span class="keyword">const</span> E1 expression1_type;
<a name="l00743"></a>00743 <span class="keyword">typedef</span> <span class="keyword">const</span> E2 expression2_type;
<a name="l00744"></a>00744 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::size_type size_type;
<a name="l00745"></a>00745 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::value_type value_type;
<a name="l00746"></a>00746
<a name="l00747"></a>00747 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00748"></a>00748 <span class="preprocessor"></span> matrix<value_type, row_major> cm (m);
<a name="l00749"></a>00749 <span class="keyword">typedef</span> <span class="keyword">typename</span> type_traits<value_type>::real_type real_type;
<a name="l00750"></a>00750 real_type merrorbound (norm_1 (m) + norm_1 (e1) * norm_1 (e2));
<a name="l00751"></a>00751 indexing_matrix_assign<scalar_plus_assign> (cm, prod (e1, e2), row_major_tag ());
<a name="l00752"></a>00752 <span class="preprocessor">#endif</span>
<a name="l00753"></a>00753 <span class="preprocessor"></span> size_type size (BOOST_UBLAS_SAME (e1 ().size2 (), e2 ().size1 ()));
<a name="l00754"></a>00754 <span class="keywordflow">for</span> (size_type k = 0; k < size; ++ k) {
<a name="l00755"></a>00755 vector<value_type> ce1 (column (e1 (), k));
<a name="l00756"></a>00756 vector<value_type> re2 (row (e2 (), k));
<a name="l00757"></a>00757 m.plus_assign (outer_prod (ce1, re2));
<a name="l00758"></a>00758 }
<a name="l00759"></a>00759 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00760"></a>00760 <span class="preprocessor"></span> BOOST_UBLAS_CHECK (norm_1 (m - cm) <= 2 * std::numeric_limits<real_type>::epsilon () * merrorbound, internal_logic ());
<a name="l00761"></a>00761 <span class="preprocessor">#endif</span>
<a name="l00762"></a>00762 <span class="preprocessor"></span> <span class="keywordflow">return</span> m;
<a name="l00763"></a>00763 }
<a name="l00764"></a>00764
<a name="l00765"></a>00765 <span class="keyword">template</span><<span class="keyword">class</span> M, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00766"></a>00766 BOOST_UBLAS_INLINE
<a name="l00767"></a>00767 M &
<a name="l00768"></a>00768 opb_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00769"></a>00769 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00770"></a>00770 M &m,
<a name="l00771"></a>00771 dense_proxy_tag, column_major_tag) {
<a name="l00772"></a>00772 <span class="keyword">typedef</span> M matrix_type;
<a name="l00773"></a>00773 <span class="keyword">typedef</span> <span class="keyword">const</span> E1 expression1_type;
<a name="l00774"></a>00774 <span class="keyword">typedef</span> <span class="keyword">const</span> E2 expression2_type;
<a name="l00775"></a>00775 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::size_type size_type;
<a name="l00776"></a>00776 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::value_type value_type;
<a name="l00777"></a>00777
<a name="l00778"></a>00778 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00779"></a>00779 <span class="preprocessor"></span> matrix<value_type, column_major> cm (m);
<a name="l00780"></a>00780 <span class="keyword">typedef</span> <span class="keyword">typename</span> type_traits<value_type>::real_type real_type;
<a name="l00781"></a>00781 real_type merrorbound (norm_1 (m) + norm_1 (e1) * norm_1 (e2));
<a name="l00782"></a>00782 indexing_matrix_assign<scalar_plus_assign> (cm, prod (e1, e2), column_major_tag ());
<a name="l00783"></a>00783 <span class="preprocessor">#endif</span>
<a name="l00784"></a>00784 <span class="preprocessor"></span> size_type size (BOOST_UBLAS_SAME (e1 ().size2 (), e2 ().size1 ()));
<a name="l00785"></a>00785 <span class="keywordflow">for</span> (size_type k = 0; k < size; ++ k) {
<a name="l00786"></a>00786 vector<value_type> ce1 (column (e1 (), k));
<a name="l00787"></a>00787 vector<value_type> re2 (row (e2 (), k));
<a name="l00788"></a>00788 m.plus_assign (outer_prod (ce1, re2));
<a name="l00789"></a>00789 }
<a name="l00790"></a>00790 <span class="preprocessor">#if BOOST_UBLAS_TYPE_CHECK</span>
<a name="l00791"></a>00791 <span class="preprocessor"></span> BOOST_UBLAS_CHECK (norm_1 (m - cm) <= 2 * std::numeric_limits<real_type>::epsilon () * merrorbound, internal_logic ());
<a name="l00792"></a>00792 <span class="preprocessor">#endif</span>
<a name="l00793"></a>00793 <span class="preprocessor"></span> <span class="keywordflow">return</span> m;
<a name="l00794"></a>00794 }
<a name="l00795"></a>00795
<a name="l00796"></a>00796 <span class="comment">// Dispatcher</span>
<a name="l00797"></a>00797
<a name="l00824"></a>00824 <span class="keyword">template</span><<span class="keyword">class</span> M, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00825"></a>00825 BOOST_UBLAS_INLINE
<a name="l00826"></a>00826 M &
<a name="l00827"></a>00827 opb_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00828"></a>00828 <span class="keyword">const</span> matrix_expression<E2> &e2,
<a name="l00829"></a>00829 M &m, <span class="keywordtype">bool</span> init = <span class="keyword">true</span>) {
<a name="l00830"></a>00830 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::value_type value_type;
<a name="l00831"></a>00831 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::storage_category storage_category;
<a name="l00832"></a>00832 <span class="keyword">typedef</span> <span class="keyword">typename</span> M::orientation_category orientation_category;
<a name="l00833"></a>00833
<a name="l00834"></a>00834 <span class="keywordflow">if</span> (init)
<a name="l00835"></a>00835 m.assign (zero_matrix<value_type> (e1 ().size1 (), e2 ().size2 ()));
<a name="l00836"></a>00836 <span class="keywordflow">return</span> opb_prod (e1, e2, m, storage_category (), orientation_category ());
<a name="l00837"></a>00837 }
<a name="l00838"></a>00838 <span class="keyword">template</span><<span class="keyword">class</span> M, <span class="keyword">class</span> E1, <span class="keyword">class</span> E2>
<a name="l00839"></a>00839 BOOST_UBLAS_INLINE
<a name="l00840"></a>00840 M
<a name="l00841"></a>00841 opb_prod (<span class="keyword">const</span> matrix_expression<E1> &e1,
<a name="l00842"></a>00842 <span class="keyword">const</span> matrix_expression<E2> &e2) {
<a name="l00843"></a>00843 <span class="keyword">typedef</span> M matrix_type;
<a name="l00844"></a>00844
<a name="l00845"></a>00845 matrix_type m (e1 ().size1 (), e2 ().size2 ());
<a name="l00846"></a>00846 <span class="keywordflow">return</span> opb_prod (e1, e2, m, <span class="keyword">true</span>);
<a name="l00847"></a>00847 }
<a name="l00848"></a>00848
<a name="l00849"></a>00849 }}}
<a name="l00850"></a>00850
<a name="l00851"></a>00851 <span class="preprocessor">#endif</span>
</pre></div></div>
<hr size="1"/><address style="text-align: right;"><small>Generated on Sun Jul 4 20:30:58 2010 for ublas by
<a href="http://www.doxygen.org/index.html">
<img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.6.1 </small></address>
</body>
</html>
