source: vis_dev/vis-2.1/share/help/dynamic_var_orderingCmd.txt @ 11

Last change on this file since 11 was 11, checked in by cecile, 13 years ago

Add vis
File size: 5.8 KB
Line 
1
2  dynamic_var_ordering - control the application of dynamic variable ordering
3     _________________________________________________________________
4
5   dynamic_var_ordering [-d] [-e <method>] [-f <method>] [-h]
6
7   Control  the application of dynamic variable ordering to the flattened
8   network.  Dynamic ordering is a technique to reorder the MDD variables
9   to  reduce  the  size  of  the  existing  MDDs.  When  no  options are
10   specified,  the  current  status  of dynamic ordering is displayed. At
11   most  one  of  the  options  -e,  -f,  and -d should be specified. The
12   commands  flatten_hierarchy  and  static_order  must be invoked before
13   this command.
14
15   Dynamic  ordering may be time consuming, but can often reduce the size
16   of  the  MDDs dramatically. The good points to invoke dynamic ordering
17   explicitly   (using   the   -f   option)   are   after   the  commands
18   build_partition_mdds  and  print_img_info. If dynamic ordering finds a
19   good  ordering,  then  you  may  wish  to  save  this  ordering (using
20   write_order)  and  reuse  it  (using static_order -s) in the future. A
21   common sequence used to get a good ordering is the following:
22
23  init_verify
24  print_img_info
25  dynamic_var_ordering -f sift
26  write_order
27  flatten_hierarchy
28  static_order -s input_and_latch -f
29  build_partition_mdds
30  print_img_info
31  dynamic_var_ordering -f sift
32 
33
34   For  many  large  examples, there is no single best variable order, or
35   that  order  is  hard  to  find. For example, the best ordering during
36   partitioning  of  the  network may be different from the best ordering
37   during  a  model check. In that case you can use automatic reordering,
38   using  the  -e option. This will trigger reordering whenever the total
39   size of the MDD increases by a certain factor. Often, the init command
40   is replaced by the following sequence:
41  flatten_hierarchy
42  static_order
43  dynamic_var_ordering -e sift
44  build_partition_mdds
45 
46
47   Command options:
48
49   -d
50          Disable dynamic ordering from triggering automatically.
51
52   -e <method>
53          Enable  dynamic  ordering  to  trigger automatically whenever a
54          certain  threshold on the overall MDD size is reached. "Method"
55          must be one of the following:
56
57          window: Permutes the variables within windows of three adjacent
58          variables  so as to minimize the overall MDD size. This process
59          is repeated until no more reduction in size occurs.
60
61          sift:  Moves  each  variable  throughout  the  order to find an
62          optimal   position   for  that  variable  (assuming  all  other
63          variables  are  fixed).  This  generally  achieves greater size
64          reductions than the window method, but is slower.
65
66          The following methods are only available if VIS has been linked
67          with the Bdd package from the University of Colorado (cuBdd).
68
69          random:  Pairs of variables are randomly chosen, and swapped in
70          the  order.  The  swap  is  performed  by  a series of swaps of
71          adjacent  variables. The best order among those obtained by the
72          series  of  swaps  is  retained. The number of pairs chosen for
73          swapping equals the number of variables in the diagram.
74
75          random_pivot:  Same as random, but the two variables are chosen
76          so that the first is above the variable with the largest number
77          of  nodes, and the second is below that variable. In case there
78          are several variables tied for the maximum number of nodes, the
79          one closest to the root is used.
80
81          sift_converge:  The  sift  method  is iterated until no further
82          improvement is obtained.
83
84          symmetry_sift:  This  method  is an implementation of symmetric
85          sifting. It is similar to sifting, with one addition: Variables
86          that become adjacent during sifting are tested for symmetry. If
87          they  are  symmetric,  they are linked in a group. Sifting then
88          continues  with  a  group  being  moved,  instead  of  a single
89          variable.
90
91          symmetry_sift_converge:  The  symmetry_sift  method is iterated
92          until no further improvement is obtained.
93
94          window{2,3,4}:  Permutes  the  variables  within  windows  of n
95          adjacent variables, where "n" can be either 2, 3 or 4, so as to
96          minimize the overall MDD size.
97
98          window{2,3,4}_converge:  The  window{2,3,4}  method is iterated
99          until no further improvement is obtained.
100
101          group_sift:  This  method is similar to symmetry_sift, but uses
102          more general criteria to create groups.
103
104          group_sift_converge: The group_sift method is iterated until no
105          further improvement is obtained.
106
107          lazy_sift:  This  method  is  similar  to  group_sift,  but the
108          creation  of  groups  takes into account the pairing of present
109          and next state variables.
110
111          annealing:  This  method  is  an  implementation  of  simulated
112          annealing  for  variable  ordering.  This method is potentially
113          very slow.
114
115          genetic:   This  method  is  an  implementation  of  a  genetic
116          algorithm  for  variable  ordering.  This method is potentially
117          very slow.
118
119   -f <method>
120          Force  dynamic  ordering  to be invoked immediately. The values
121          for method are the same as in option -e.
122
123   -h
124          Print the command usage.
125
126   -v <#>
127          Verbosity  level.  Default  is  0.  For  values greater than 0,
128          statistics  pertaining  to  reordering will be printed whenever
129          reordering occurs.
130     _________________________________________________________________
131
132   Last updated on 20050519 10h16
Note: See TracBrowser for help on using the repository browser.