Changeset 1525


Ignore:
Timestamp:
Sep 22, 2009 6:25:12 AM (11 years ago)
Author:
bradbell
Message:

trunk: Change description of connection to description of vector of sets.

Location:
trunk/cppad/local
Files:
9 edited

Legend:

Unmodified
Added
Removed
  • trunk/cppad/local/cond_op.hpp

    r1524 r1525  
    504504
    505505\param hes_sparsity
    506 The connections in \a hes_sparsity for the from node with index \a i_z
    507 are the Hessian sparsity pattern for the function G
     506The set with index \a i_z in \a hes_sparsity
     507is the Hessian sparsity pattern for the function G
    508508where one of the partials is with respect to z.
    509509\n
    510510\n
    511511If y_2 is a variable,
    512 the connections in \a hes_sparsity for the from node with index \a arg[4]
    513 are the Hessian sparsity pattern
     512the set with index \a arg[4] in \a hes_sparsity
     513is the Hessian sparsity pattern
    514514where one of the partials is with respect to y_2.
    515515On input, this pattern corresponds to the function G.
     
    518518\n
    519519If y_3 is a variable,
    520 the connections in \a hes_sparsity for the from node with index \a arg[5]
    521 are the Hessian sparsity pattern
     520the set with index \a arg[5] in \a hes_sparsity
     521is the Hessian sparsity pattern
    522522where one of the partials is with respect to y_3.
    523523On input, this pattern corresponds to the function G.
  • trunk/cppad/local/load_op.hpp

    r1524 r1525  
    350350Reverse mode Hessian sparsity operations for LdpOp and LdvOp
    351351
    352 This routine is given the (EDIT THIS) connections corresponding to
     352This routine is given the sparsity patterns for
    353353G(z , v[x] , w , u ... )
    354 and it uses them to compute the (EDIT THIS) connections corresponding to
     354and it uses them to compute the sparsity patterns for
    355355\verbatim
    356356        H( v[x] , w , u , ... ) = G[ z( v[x] ) , v[x] , w , u , ... ]
  • trunk/cppad/local/prototype_op.hpp

    r1524 r1525  
    10581058\n
    10591059\a combined[ \a arg[0] - 1 ]
    1060 is the from node index for the vector v  in the \a vecad_sparsity
    1061 connection object.
     1060is the index of the set corresponding to the vector v  in \a vecad_sparsity.
    10621061We use the notation i_v for this value; i.e.,
    10631062\verbatim
     
    10661065
    10671066\param var_sparsity
    1068 The connections (EDIT THIS) in \a var_sparsity with from node index \a i_z
    1069 are the sparsity pattern for z.
    1070 These connections (EDIT THIS) are an output for forward mode operations,
     1067The set with index \a i_z in \a var_sparsity is the sparsity pattern for z.
     1068This is an output for forward mode operations,
    10711069and an input for reverse mode operations.
    10721070
    10731071\param vecad_sparsity
    1074 The connections (EDIT THIS) in \a vecad_sparsity with from node index \a i_v
    1075 are the sparsity bit pattern for the vector v.
    1076 These connections (EDIT THIS) are an input for forward mode operations.
     1072The set with index \a i_v is the sparsity pattern for the vector v.
     1073This is an input for forward mode operations.
    10771074For reverse mode operations,
    10781075the sparsity pattern for z is added to the sparsity pattern for v.
     
    12331230\a arg[2]
    12341231\n
    1235 index corresponding to the third operand for this operator;
    1236 It is lso the from index for y in the \a var_sparsity connection object.
     1232The set with index \a arg[2] in \a var_sparsity
     1233is the sparsity pattern corresponding to y.
    12371234(Note that \a arg[2] > 0 because y is a variable.)
    12381235
     
    12421239\param combined
    12431240\a combined [ arg[0] - 1 ]
    1244 is the from index for the VecAD vector v in the \a vecad_sparsity
    1245 connection object.
     1241is the index of the set in \a vecad_sparsity corresponding
     1242to the sparsity pattern for the vector v.
    12461243We use the notation i_v below which is defined by
    12471244\verbatim
     
    12501247
    12511248\param var_sparsity
    1252 The connections (EDIT THIS) in \a var_sparsity corresponding to the from index \a arg[2]
    1253 correspond to the variable y.
    1254 These are an input for forward mode operations.
     1249The set  with index \a arg[2] in \a var_sparsity
     1250is the sparsity pattern for y.
     1251This is an input for forward mode operations.
    12551252For reverse mode operations:
    12561253The sparsity pattern for v is added to the spartisy pattern for y.
    12571254
    12581255\param vecad_sparsity
    1259 The connections (EDIT THIS) in \a vecad_sparsity corresponding to the from index \a i_v
    1260 correspond to the vector v.
    1261 These are an input for reverse mode operations.
     1256The set with index \a i_v in \a vecad_sparsity
     1257is the sparsity pattern for v.
     1258This is an input for reverse mode operations.
    12621259For forward mode operations, the sparsity pattern for y is added
    12631260to the sparsity pattern for the vector v.
     
    13171314
    13181315\param for_jac_sparsity
    1319 The connections (EDIT THIS) for the from node \a i_x
     1316The set with index \a i_x in for_jac_sparsity
    13201317is the forward mode Jacobian sparsity pattern for the variable x.
    13211318
    13221319\param rev_hes_sparsity
    1323 The connections (EDIT THIS) for the from node with from index \a i_z in \a rev_hes_sparsity
     1320The set with index \a i_z in in \a rev_hes_sparsity
    13241321is the Hessian sparsity pattern for the fucntion G
    13251322where one of the partials derivative is with respect to z.
    13261323\n
    13271324\n
    1328 The connections (EDIT THIS) for the form node with index \a i_x in \a rev_hes_sparsity
     1325The set with index \a i_x in \a rev_hes_sparsity
    13291326is the Hessian sparsity pattern
    13301327where one of the partials derivative is with respect to x.
     
    13891386
    13901387\param for_jac_sparsity
    1391 The connections (EDIT THIS) in \a for_jac_sparsity for the
    1392 from node with index \a arg[0] are the forward sparsity pattern for x.
    1393 \n
    1394 \n
    1395 The connections (EDIT THIS) in \a for_jac_sparsity for the
    1396 from node with index \a arg[1] are the forward sparsity pattern for y.
     1388The set with index \a arg[0] in \a for_jac_sparsity for the
     1389is the forward Jacobian sparsity pattern for x.
     1390\n
     1391\n
     1392The set with index \a arg[1] in \a for_jac_sparsity
     1393is the forward sparsity pattern for y.
    13971394
    13981395\param rev_hes_sparsity
    1399 The connections (EDIT THIS) in \a rev_hes_sparsity for the
    1400 from node with index \a i_z are the Hessian sparsity pattern
    1401 for the function G
     1396The set wiht index \a i_x in \a rev_hes_sparsity
     1397is the Hessian sparsity pattern for the function G
    14021398where one of the partial derivatives is with respect to z.
    14031399\n
    14041400\n
    1405 The connections (EDIT THIS) in \a rev_hes_sparsity for the
    1406 from node wiht index \a arg[0] are the Hessian sparsity pattern
    1407 where one of the partial derivatives is with respect to x.
     1401The set with index \a arg[0] in  \a rev_hes_sparsity
     1402is the Hessian sparsity pattern where one of the
     1403partial derivatives is with respect to x.
    14081404On input, it corresponds to the function G,
    14091405and on output it correspondst to H.
    14101406\n
    14111407\n
    1412 The connections (EDIT THIS) in \a rev_hes_sparsity for the
    1413 from node wiht index \a arg[1] are the Hessian sparsity pattern
    1414 where one of the partial derivatives is with respect to y.
     1408The set with index \a arg[1] in \a rev_hes_sparsity
     1409is the Hessian sparsity pattern where one of the
     1410partial derivatives is with respect to y.
    14151411On input, it corresponds to the function G,
    14161412and on output it correspondst to H.
  • trunk/cppad/local/rev_hes_sweep.hpp

    r1524 r1525  
    8989\b Output: For j = 1 , ... , \a n,
    9090the reverse Hessian sparsity pattern for the independent dependent variable
    91 with index (j-1) is given by the from connections (EDIT THIS) for the node with index j
     91with index (j-1) is given by the set with index j
    9292in \a rev_hes_sparse.
    9393The values in the rest of \a rev_hes_sparse are not specified; i.e.,
     
    122122        CPPAD_ASSERT_UNKNOWN( numvar > 0 );
    123123
    124         // number of to nodes in (EDIT THIS) connections
     124        // upper limit exclusive for set elements
    125125        size_t limit   = rev_hes_sparse.limit();
    126126        CPPAD_ASSERT_UNKNOWN( rev_hes_sparse.limit() == limit );
  • trunk/cppad/local/rev_jac_sweep.hpp

    r1524 r1525  
    7474\b Output: For j = 1 , ... , \a n,
    7575the sparsity pattern for the dependent variable with index (j-1)
    76 is given by the connections for the from node with index j in
    77 \a var_sparsity.
     76is given by the set with index index j in \a var_sparsity.
    7877*/
    7978
     
    102101        CPPAD_ASSERT_UNKNOWN( var_sparsity.n_set() == numvar );
    103102
    104         // number of to nodes in the connection
    105         size_t n_to = var_sparsity.limit();
     103        // upper limit (exclusive) for elements in the set
     104        size_t limit = var_sparsity.limit();
    106105
    107106        // vecad_sparsity contains a sparsity pattern for each VecAD object.
     
    111110        size_t num_vecad_vec   = play->num_rec_vecad_vec();
    112111        vector_pack      vecad_sparsity;
    113         vecad_sparsity.resize(num_vecad_vec, n_to);
     112        vecad_sparsity.resize(num_vecad_vec, limit);
    114113        size_t* vecad_ind      = CPPAD_NULL;
    115114        if( num_vecad_vec > 0 )
     
    135134# if CPPAD_REV_JAC_SWEEP_TRACE
    136135        std::cout << std::endl;
    137         CppAD::vector<bool> z_value(n_to);
     136        CppAD::vector<bool> z_value(limit);
    138137# endif
    139138        while(i_op > 1)
     
    145144
    146145# if CPPAD_REV_JAC_SWEEP_TRACE
    147                 for(j = 0; j < n_to; j++)
     146                for(j = 0; j < limit; j++)
    148147                        z_value[j] = var_sparsity.get_element(i_var, j);
    149148                printOp(
  • trunk/cppad/local/rev_sparse_hes.hpp

    r1524 r1525  
    223223        RevJac       = CPPAD_TRACK_NEW_VEC(total_num_var_, RevJac);     
    224224
    225         // connection object that will hold packed reverse Hessain values
     225        // vector of sets that will hold packed reverse Hessain values
    226226        vector_pack      rev_hes_sparsity;
    227227        rev_hes_sparsity.resize(total_num_var_, q);
  • trunk/cppad/local/rev_sparse_jac.hpp

    r1524 r1525  
    178178        );
    179179
    180         // connection object that will hold the results
     180        // vector of sets that will hold the results
    181181        vector_pack      var_sparsity;
    182182        var_sparsity.resize(total_num_var_, p);
    183183
    184         // set from node connections corresponding to dependent variables
     184        // The sparsity pattern corresponding to the dependent variables
    185185        for(i = 0; i < m; i++)
    186186        {       CPPAD_ASSERT_UNKNOWN( dep_taddr_[i] < total_num_var_ );
  • trunk/cppad/local/sparse_binary_op.hpp

    r1524 r1525  
    128128
    129129\param sparsity
    130 The (EDIT THIS) connections in \a sparsity for the
    131 from node with index \a i_z are the sparsity pattern for z
    132 and corresponding ot the function G.
    133 \n
    134 \n
    135 The (EDIT THIS) connections in \a sparsity for the
    136 from node with index \a arg[0] are the sparsity pattern for x.
    137 On input, they correspond to the function G,
    138 and on output they correspond to H.
    139 \n
    140 \n
    141 The (EDIT THIS) connections in \a sparsity for the
    142 from node with index \a arg[1] are the sparsity pattern for y.
    143 On input, they correspond to the function G,
    144 and on output they correspond to H.
     130The set with index \a i_z in \a sparsity
     131is the sparsity pattern for z corresponding ot the function G.
     132\n
     133\n
     134The set with index \a arg[0] in \a sparsity
     135is the sparsity pattern for x.
     136On input, it corresponds to the function G,
     137and on output it corresponds to H.
     138\n
     139\n
     140The set with index \a arg[1] in \a sparsity
     141is the sparsity pattern for y.
     142On input, it corresponds to the function G,
     143and on output it corresponds to H.
     144\n
     145\n
    145146
    146147\par Checked Assertions:
  • trunk/cppad/local/store_op.hpp

    r1524 r1525  
    185185Reverse mode sparsity operations for StpvOp and StvvOp
    186186
    187 This routine is given the (EDIT THIS) connections corresponding to
    188 G(v[x], y , w , u ... )
    189 and it uses them to compute the partial derivatives of
     187This routine is given the sparsity patterns for
     188G(v[x], y , w , u ... ) and it uses them to compute the
     189sparsity patterns for
    190190\verbatim
    191191        H(y , w , u , ... ) = G[ v[x], y , w , u , ... ]
     
    218218Reverse mode sparsity operations for StpvOp and StvvOp
    219219
    220 This routine is given the (EDIT THIS) connections corresponding to
     220This routine is given the sparsity patterns for
    221221G(v[x], y , w , u ... )
    222 and it uses them to compute the partial derivatives of
     222and it uses them to compute the sparsity patterns for
    223223\verbatim
    224224        H(y , w , u , ... ) = G[ v[x], y , w , u , ... ]
Note: See TracChangeset for help on using the changeset viewer.