doc/named-param.dox
author Peter Kovacs <kpeter@inf.elte.hu>
Tue, 15 Mar 2011 19:32:21 +0100
changeset 936 ddd3c0d3d9bf
parent 269 f5965bbf1353
child 1142 2f479109a71d
permissions -rw-r--r--
Implement the scaling Price Refinement heuristic in CostScaling (#417)
instead of Early Termination.

These two heuristics are similar, but the newer one is faster
and not only makes it possible to skip some epsilon phases, but
it can improve the performance of the other phases, as well.
kpeter@269
     1
/* -*- mode: C++; indent-tabs-mode: nil; -*-
alpar@267
     2
 *
kpeter@269
     3
 * This file is a part of LEMON, a generic C++ optimization library.
alpar@267
     4
 *
alpar@440
     5
 * Copyright (C) 2003-2009
alpar@267
     6
 * Egervary Jeno Kombinatorikus Optimalizalasi Kutatocsoport
alpar@267
     7
 * (Egervary Research Group on Combinatorial Optimization, EGRES).
alpar@267
     8
 *
alpar@267
     9
 * Permission to use, modify and distribute this software is granted
alpar@267
    10
 * provided that this copyright notice appears in all copies. For
alpar@267
    11
 * precise terms see the accompanying LICENSE file.
alpar@267
    12
 *
alpar@267
    13
 * This software is provided "AS IS" with no warranty of any kind,
alpar@267
    14
 * express or implied, and with no claim as to its suitability for any
alpar@267
    15
 * purpose.
alpar@267
    16
 *
alpar@267
    17
 */
alpar@267
    18
alpar@267
    19
/*!
alpar@267
    20
alpar@267
    21
\page named-param Named Parameters
alpar@267
    22
alpar@267
    23
\section named-func-param Named Function Parameters
alpar@267
    24
alpar@268
    25
Several modern languages provide a convenient way to refer the
alpar@268
    26
function parameters by name also when you call the function. It is
alpar@268
    27
especially comfortable in case of a function having tons of parameters
kpeter@269
    28
with natural default values. Sadly, C++ lack this amenity.
alpar@268
    29
alpar@268
    30
However, with a crafty trick and with some little
alpar@268
    31
inconvenience, it is possible to emulate is.
alpar@268
    32
The example below shows how to do it.
alpar@267
    33
alpar@267
    34
\code
kpeter@269
    35
class namedFn
alpar@267
    36
{
alpar@267
    37
  int _id;
alpar@267
    38
  double _val;
alpar@267
    39
  int _dim;
kpeter@269
    40
alpar@267
    41
  public:
alpar@267
    42
  namedFn() : _id(0), _val(1), _dim(2) {}
alpar@267
    43
  namedFn& id(int p)     { _id  = p ; return *this; }
alpar@267
    44
  namedFn& val(double p) { _val = p ; return *this; }
alpar@267
    45
  namedFn& dim(int p)    { _dim = p ; return *this; }
alpar@267
    46
alpar@267
    47
  run() {
kpeter@269
    48
    std::cout << "Here comes the function itself\n" <<
kpeter@269
    49
              << "With parameters "
kpeter@269
    50
              << _id << ", " << _val << ", " << _dim << std::endl;
alpar@267
    51
  }
alpar@267
    52
};
alpar@267
    53
\endcode
alpar@267
    54
alpar@268
    55
Then you can use it like this.
alpar@267
    56
alpar@267
    57
\code
alpar@267
    58
namedFn().id(3).val(2).run();
alpar@267
    59
\endcode
alpar@267
    60
alpar@268
    61
The trick is obvious, each "named parameter" changes one component of
alpar@268
    62
the underlying class, then gives back a reference to it. Finally,
alpar@268
    63
<tt>run()</tt> executes the algorithm itself.
alpar@268
    64
alpar@267
    65
\note Although it is a class, namedFn is used pretty much like as it were
alpar@268
    66
a function. That it why we called it namedFn instead of \c NamedFn.
alpar@267
    67
alpar@268
    68
\note In fact, the final <tt>.run()</tt> could be made unnecessary,
alpar@268
    69
because the algorithm could also be implemented in the destructor of
alpar@268
    70
\c namedFn instead. This however would make it impossible to implement
alpar@268
    71
functions with return values, and would also cause serious problems when
alpar@268
    72
implementing \ref named-templ-func-param "named template parameters".
alpar@268
    73
<b>Therefore, by convention, <tt>.run()</tt> must be used
alpar@268
    74
explicitly to execute a function having named parameters
alpar@268
    75
everywhere in LEMON.</b>
alpar@267
    76
alpar@268
    77
\section named-templ-func-param Named Function Template Parameters
alpar@268
    78
kpeter@269
    79
A named parameter can also be a template function. The usage is
alpar@268
    80
exactly the same, but the implementation behind is a kind of black
alpar@268
    81
magic and they are the dirtiest part of the LEMON code.
alpar@268
    82
alpar@268
    83
You will probably never need to know how it works, but if you really
alpar@268
    84
committed, have a look at \ref lemon/graph_to_eps.h for an example.
alpar@267
    85
alpar@267
    86
\section traits-classes Traits Classes
alpar@267
    87
alpar@268
    88
A similar game can also be played when defining classes. In this case
alpar@268
    89
the type of the class attributes can be changed. Initially we have to
alpar@268
    90
define a special class called <em>Traits Class</em> defining the
alpar@268
    91
default type of the attributes. Then the types of these attributes can
alpar@268
    92
be changed in the same way as described in the next section.
alpar@268
    93
alpar@268
    94
See \ref lemon::DijkstraDefaultTraits for an
alpar@268
    95
example how a traits class implementation looks like.
alpar@267
    96
alpar@267
    97
\section named-templ-param Named Class Template Parameters
alpar@267
    98
alpar@267
    99
If we would like to change the type of an attribute in a class that
alpar@267
   100
was instantiated by using a traits class as a template parameter, and
alpar@268
   101
the class contains named parameters, we do not have to instantiate again
alpar@268
   102
the class with new traits class, but instead adaptor classes can
alpar@268
   103
be used as shown in the following example.
alpar@267
   104
alpar@267
   105
\code
kpeter@269
   106
Dijkstra<>::SetPredMap<NullMap<Node,Arc> >::Create
alpar@267
   107
\endcode
alpar@267
   108
alpar@267
   109
It can also be used in conjunction with other named template
alpar@267
   110
parameters in arbitrary order.
alpar@267
   111
alpar@267
   112
\code
kpeter@269
   113
Dijkstra<>::SetDistMap<MyMap>::SetPredMap<NullMap<Node,Arc> >::Create
alpar@267
   114
\endcode
alpar@267
   115
alpar@267
   116
The result will be an instantiated Dijkstra class, in which the
alpar@267
   117
DistMap and the PredMap is modified.
alpar@267
   118
alpar@267
   119
*/