doc/glpk03.tex
author Alpar Juttner <alpar@cs.elte.hu>
Mon, 06 Dec 2010 13:09:21 +0100
changeset 1 c445c931472f
permissions -rw-r--r--
Import glpk-4.45

- Generated files and doc/notes are removed
alpar@1
     1
%* glpk03.tex *%
alpar@1
     2
alpar@1
     3
\chapter{Utility API routines}
alpar@1
     4
alpar@1
     5
\section{Problem data reading/writing routines}
alpar@1
     6
alpar@1
     7
\subsection{glp\_read\_mps---read problem data in MPS format}
alpar@1
     8
alpar@1
     9
\subsubsection*{Synopsis}
alpar@1
    10
alpar@1
    11
\begin{verbatim}
alpar@1
    12
int glp_read_mps(glp_prob *lp, int fmt, const void *parm,
alpar@1
    13
      const char *fname);
alpar@1
    14
\end{verbatim}
alpar@1
    15
alpar@1
    16
\subsubsection*{Description}
alpar@1
    17
alpar@1
    18
The routine \verb|glp_read_mps| reads problem data in MPS format from a
alpar@1
    19
text file. (The MPS format is described in Appendix \ref{champs}, page
alpar@1
    20
\pageref{champs}.)
alpar@1
    21
alpar@1
    22
The parameter \verb|fmt| specifies the MPS format version as follows:
alpar@1
    23
alpar@1
    24
\begin{tabular}{@{}ll}
alpar@1
    25
\verb|GLP_MPS_DECK| & fixed (ancient) MPS format; \\
alpar@1
    26
\verb|GLP_MPS_FILE| & free (modern) MPS format. \\
alpar@1
    27
\end{tabular}
alpar@1
    28
alpar@1
    29
The parameter \verb|parm| is reserved for use in the future and must be
alpar@1
    30
specified as \verb|NULL|.
alpar@1
    31
alpar@1
    32
The character string \verb|fname| specifies a name of the text file to
alpar@1
    33
be read in. (If the file name ends with suffix `\verb|.gz|', the file is
alpar@1
    34
assumed to be compressed, in which case the routine \verb|glp_read_mps|
alpar@1
    35
decompresses it ``on the fly''.)
alpar@1
    36
alpar@1
    37
Note that before reading data the current content of the problem object
alpar@1
    38
is completely erased with the routine \verb|glp_erase_prob|.
alpar@1
    39
alpar@1
    40
\subsubsection*{Returns}
alpar@1
    41
alpar@1
    42
If the operation was successful, the routine \verb|glp_read_mps|
alpar@1
    43
returns zero. Otherwise, it prints an error message and returns
alpar@1
    44
non-zero.
alpar@1
    45
alpar@1
    46
\subsection{glp\_write\_mps---write problem data in MPS format}
alpar@1
    47
alpar@1
    48
\subsubsection*{Synopsis}
alpar@1
    49
alpar@1
    50
\begin{verbatim}
alpar@1
    51
int glp_write_mps(glp_prob *lp, int fmt, const void *parm,
alpar@1
    52
      const char *fname);
alpar@1
    53
\end{verbatim}
alpar@1
    54
alpar@1
    55
\subsubsection*{Description}
alpar@1
    56
alpar@1
    57
The routine \verb|glp_write_mps| writes problem data in MPS format to a
alpar@1
    58
text file. (The MPS format is described in Appendix \ref{champs}, page
alpar@1
    59
\pageref{champs}.)
alpar@1
    60
alpar@1
    61
The parameter \verb|fmt| specifies the MPS format version as follows:
alpar@1
    62
alpar@1
    63
\begin{tabular}{@{}ll}
alpar@1
    64
\verb|GLP_MPS_DECK| & fixed (ancient) MPS format; \\
alpar@1
    65
\verb|GLP_MPS_FILE| & free (modern) MPS format. \\
alpar@1
    66
\end{tabular}
alpar@1
    67
alpar@1
    68
The parameter \verb|parm| is reserved for use in the future and must be
alpar@1
    69
specified as \verb|NULL|.
alpar@1
    70
alpar@1
    71
The character string \verb|fname| specifies a name of the text file to
alpar@1
    72
be written out. (If the file name ends with suffix `\verb|.gz|', the
alpar@1
    73
file is assumed to be compressed, in which case the routine
alpar@1
    74
\verb|glp_write_mps| performs automatic compression on writing it.)
alpar@1
    75
alpar@1
    76
\subsubsection*{Returns}
alpar@1
    77
alpar@1
    78
If the operation was successful, the routine \verb|glp_write_mps|
alpar@1
    79
returns zero. Otherwise, it prints an error message and returns
alpar@1
    80
non-zero.
alpar@1
    81
alpar@1
    82
\subsection{glp\_read\_lp---read problem data in CPLEX LP format}
alpar@1
    83
alpar@1
    84
\subsubsection*{Synopsis}
alpar@1
    85
alpar@1
    86
\begin{verbatim}
alpar@1
    87
int glp_read_lp(glp_prob *lp, const void *parm,
alpar@1
    88
      const char *fname);
alpar@1
    89
\end{verbatim}
alpar@1
    90
alpar@1
    91
\subsubsection*{Description}
alpar@1
    92
alpar@1
    93
The routine \verb|glp_read_lp| reads problem data in CPLEX LP format
alpar@1
    94
from a text file. (The CPLEX LP format is described in Appendix
alpar@1
    95
\ref{chacplex}, page \pageref{chacplex}.)
alpar@1
    96
alpar@1
    97
The parameter \verb|parm| is reserved for use in the future and must be
alpar@1
    98
specified as \verb|NULL|.
alpar@1
    99
alpar@1
   100
The character string \verb|fname| specifies a name of the text file to
alpar@1
   101
be read in. (If the file name ends with suffix `\verb|.gz|', the file is
alpar@1
   102
assumed to be compressed, in which case the routine \verb|glp_read_lp|
alpar@1
   103
decompresses it ``on the fly''.)
alpar@1
   104
alpar@1
   105
Note that before reading data the current content of the problem object
alpar@1
   106
is completely erased with the routine \verb|glp_erase_prob|.
alpar@1
   107
alpar@1
   108
\subsubsection*{Returns}
alpar@1
   109
alpar@1
   110
If the operation was successful, the routine \verb|glp_read_lp| returns
alpar@1
   111
zero. Otherwise, it prints an error message and returns non-zero.
alpar@1
   112
alpar@1
   113
\subsection{glp\_write\_lp---write problem data in CPLEX LP format}
alpar@1
   114
alpar@1
   115
\subsubsection*{Synopsis}
alpar@1
   116
alpar@1
   117
\begin{verbatim}
alpar@1
   118
int glp_write_lp(glp_prob *lp, const void *parm,
alpar@1
   119
      const char *fname);
alpar@1
   120
\end{verbatim}
alpar@1
   121
alpar@1
   122
\subsubsection*{Description}
alpar@1
   123
alpar@1
   124
The routine \verb|glp_write_lp| writes problem data in CPLEX LP format
alpar@1
   125
to a text file. (The CPLEX LP format is described in Appendix
alpar@1
   126
\ref{chacplex}, page \pageref{chacplex}.)
alpar@1
   127
alpar@1
   128
The parameter \verb|parm| is reserved for use in the future and must be
alpar@1
   129
specified as \verb|NULL|.
alpar@1
   130
alpar@1
   131
The character string \verb|fname| specifies a name of the text file to
alpar@1
   132
be written out. (If the file name ends with suffix `\verb|.gz|', the
alpar@1
   133
file is assumed to be compressed, in which case the routine
alpar@1
   134
\verb|glp_write_lp| performs automatic compression on writing it.)
alpar@1
   135
alpar@1
   136
\subsubsection*{Returns}
alpar@1
   137
alpar@1
   138
If the operation was successful, the routine \verb|glp_write_lp|
alpar@1
   139
returns zero. Otherwise, it prints an error message and returns
alpar@1
   140
non-zero.
alpar@1
   141
alpar@1
   142
\subsection{glp\_read\_prob---read problem data in GLPK format}
alpar@1
   143
alpar@1
   144
\subsubsection*{Synopsis}
alpar@1
   145
alpar@1
   146
\begin{verbatim}
alpar@1
   147
int glp_read_prob(glp_prob *P, int flags, const char *fname);
alpar@1
   148
\end{verbatim}
alpar@1
   149
alpar@1
   150
\subsubsection*{Description}
alpar@1
   151
alpar@1
   152
The routine \verb|glp_read_prob| reads problem data in the GLPK LP/MIP
alpar@1
   153
format from a text file. (For description of the GLPK LP/MIP format see
alpar@1
   154
below.)
alpar@1
   155
alpar@1
   156
The parameter \verb|flags| is reserved for use in the future and should
alpar@1
   157
be specified as zero.
alpar@1
   158
alpar@1
   159
The character string \verb|fname| specifies a name of the text file to
alpar@1
   160
be read in. (If the file name ends with suffix `\verb|.gz|', the file
alpar@1
   161
is assumed to be compressed, in which case the routine
alpar@1
   162
\verb|glp_read_prob| decompresses it ``on the fly''.)
alpar@1
   163
alpar@1
   164
Note that before reading data the current content of the problem object
alpar@1
   165
is completely erased with the routine \verb|glp_erase_prob|.
alpar@1
   166
alpar@1
   167
\subsubsection*{Returns}
alpar@1
   168
alpar@1
   169
If the operation was successful, the routine \verb|glp_read_prob|
alpar@1
   170
returns zero. Otherwise, it prints an error message and returns
alpar@1
   171
non-zero.
alpar@1
   172
alpar@1
   173
\subsubsection*{GLPK LP/MIP format}
alpar@1
   174
alpar@1
   175
The GLPK LP/MIP format is a DIMACS-like format.\footnote{The DIMACS
alpar@1
   176
formats were developed by the Center for Discrete Mathematics and
alpar@1
   177
Theoretical Computer Science (DIMACS) to facilitate exchange of problem
alpar@1
   178
data. For details see: {\tt <http://dimacs.rutgers.edu/Challenges/>}. }
alpar@1
   179
The file in this format is a plain ASCII text file containing lines of
alpar@1
   180
several types described below. A line is terminated with the end-of-line
alpar@1
   181
character. Fields in each line are separated by at least one blank
alpar@1
   182
space. Each line begins with a one-character designator to identify the
alpar@1
   183
line type.
alpar@1
   184
alpar@1
   185
The first line of the data file must be the problem line (except
alpar@1
   186
optional comment lines, which may precede the problem line). The last
alpar@1
   187
line of the data file must be the end line. Other lines may follow in
alpar@1
   188
arbitrary order, however, duplicate lines are not allowed.
alpar@1
   189
alpar@1
   190
\paragraph{Comment lines.} Comment lines give human-readable
alpar@1
   191
information about the data file and are ignored by GLPK routines.
alpar@1
   192
Comment lines can appear anywhere in the data file. Each comment line
alpar@1
   193
begins with the lower-case character \verb|c|.
alpar@1
   194
alpar@1
   195
\begin{verbatim}
alpar@1
   196
   c This is an example of comment line
alpar@1
   197
\end{verbatim}
alpar@1
   198
alpar@1
   199
\paragraph{Problem line.} There must be exactly one problem line in the
alpar@1
   200
data file. This line must appear before any other lines except comment
alpar@1
   201
lines and has the following format:
alpar@1
   202
alpar@1
   203
\begin{verbatim}
alpar@1
   204
   p CLASS DIR ROWS COLS NONZ
alpar@1
   205
\end{verbatim}
alpar@1
   206
alpar@1
   207
The lower-case letter \verb|p| specifies that this is the problem line.
alpar@1
   208
alpar@1
   209
The \verb|CLASS| field defines the problem class and can contain either
alpar@1
   210
the keyword \verb|lp| (that means linear programming problem) or
alpar@1
   211
\verb|mip| (that means mixed integer programming problem).
alpar@1
   212
alpar@1
   213
The \verb|DIR| field defines the optimization direction (that is, the
alpar@1
   214
objective function sense) and can contain either the keyword \verb|min|
alpar@1
   215
(that means minimization) or \verb|max| (that means maximization).
alpar@1
   216
alpar@1
   217
The \verb|ROWS|, \verb|COLS|, and \verb|NONZ| fields contain
alpar@1
   218
non-negative integer values specifying, respectively, the number of
alpar@1
   219
rows (constraints), columns (variables), and non-zero constraint
alpar@1
   220
coefficients in the problem instance. Note that \verb|NONZ| value does
alpar@1
   221
not account objective coefficients.
alpar@1
   222
alpar@1
   223
\paragraph{Row descriptors.} There must be at most one row descriptor
alpar@1
   224
line in the data file for each row (constraint). This line has one of
alpar@1
   225
the following formats:
alpar@1
   226
alpar@1
   227
\begin{verbatim}
alpar@1
   228
   i ROW f
alpar@1
   229
   i ROW l RHS
alpar@1
   230
   i ROW u RHS
alpar@1
   231
   i ROW d RHS1 RHS2
alpar@1
   232
   i ROW s RHS
alpar@1
   233
\end{verbatim}
alpar@1
   234
alpar@1
   235
The lower-case letter \verb|i| specifies that this is the row
alpar@1
   236
descriptor line.
alpar@1
   237
alpar@1
   238
The \verb|ROW| field specifies the row ordinal number, an integer
alpar@1
   239
between 1 and $m$, where $m$ is the number of rows in the problem
alpar@1
   240
instance.
alpar@1
   241
alpar@1
   242
The next lower-case letter specifies the row type as follows:
alpar@1
   243
alpar@1
   244
\verb|f| --- free (unbounded) row: $-\infty<\sum a_jx_j<+\infty$;
alpar@1
   245
alpar@1
   246
\verb|l| --- inequality constraint of `$\geq$' type:
alpar@1
   247
$\sum a_jx_j\geq b$;
alpar@1
   248
alpar@1
   249
\verb|u| --- inequality constraint of `$\leq$' type:
alpar@1
   250
$\sum a_jx_j\leq b$;
alpar@1
   251
alpar@1
   252
\verb|d| --- double-sided inequality constraint:
alpar@1
   253
$b_1\leq\sum a_jx_j\leq b_2$;
alpar@1
   254
alpar@1
   255
\verb|s| --- equality constraint: $\sum a_jx_j=b$.
alpar@1
   256
alpar@1
   257
The \verb|RHS| field contains a floaing-point value specifying the
alpar@1
   258
row right-hand side. The \verb|RHS1| and \verb|RHS2| fields contain
alpar@1
   259
floating-point values specifying, respectively, the lower and upper
alpar@1
   260
right-hand sides for the double-sided row.
alpar@1
   261
alpar@1
   262
If for some row its descriptor line does not appear in the data file,
alpar@1
   263
by default that row is assumed to be an equality constraint with zero
alpar@1
   264
right-hand side.
alpar@1
   265
alpar@1
   266
\paragraph{Column descriptors.} There must be at most one column
alpar@1
   267
descriptor line in the data file for each column (variable). This line
alpar@1
   268
has one of the following formats depending on the problem class
alpar@1
   269
specified in the problem line:
alpar@1
   270
alpar@1
   271
\bigskip
alpar@1
   272
alpar@1
   273
\begin{tabular}{@{}l@{\hspace*{40pt}}l}
alpar@1
   274
LP class & MIP class \\
alpar@1
   275
\hline
alpar@1
   276
\verb|j COL f|           & \verb|j COL KIND f|           \\
alpar@1
   277
\verb|j COL l BND|       & \verb|j COL KIND l BND|       \\
alpar@1
   278
\verb|j COL u BND|       & \verb|j COL KIND u BND|       \\
alpar@1
   279
\verb|j COL d BND1 BND2| & \verb|j COL KIND d BND1 BND2| \\
alpar@1
   280
\verb|j COL s BND|       & \verb|j COL KIND s BND|       \\
alpar@1
   281
\end{tabular}
alpar@1
   282
alpar@1
   283
\bigskip
alpar@1
   284
alpar@1
   285
The lower-case letter \verb|j| specifies that this is the column
alpar@1
   286
descriptor line.
alpar@1
   287
alpar@1
   288
The \verb|COL| field specifies the column ordinal number, an integer
alpar@1
   289
between 1 and $n$, where $n$ is the number of columns in the problem
alpar@1
   290
instance.
alpar@1
   291
alpar@1
   292
The \verb|KIND| field is used only for MIP problems and specifies the
alpar@1
   293
column kind as follows:
alpar@1
   294
alpar@1
   295
\verb|c| --- continuous column;
alpar@1
   296
alpar@1
   297
\verb|i| --- integer column;
alpar@1
   298
alpar@1
   299
\verb|b| --- binary column (in this case all remaining fields must be
alpar@1
   300
omitted).
alpar@1
   301
alpar@1
   302
The next lower-case letter specifies the column type as follows:
alpar@1
   303
alpar@1
   304
\verb|f| --- free (unbounded) column: $-\infty<x<+\infty$;
alpar@1
   305
alpar@1
   306
\verb|l| --- column with lower bound: $x\geq l$;
alpar@1
   307
alpar@1
   308
\verb|u| --- column with upper bound: $x\leq u$;
alpar@1
   309
alpar@1
   310
\verb|d| --- double-bounded column: $l\leq x\leq u$;
alpar@1
   311
alpar@1
   312
\verb|s| --- fixed column: $x=s$.
alpar@1
   313
alpar@1
   314
The \verb|BND| field contains a floating-point value that specifies the
alpar@1
   315
column bound. The \verb|BND1| and \verb|BND2| fields contain
alpar@1
   316
floating-point values specifying, respectively, the lower and upper
alpar@1
   317
bounds for the double-bounded column.
alpar@1
   318
alpar@1
   319
If for some column its descriptor line does not appear in the file, by
alpar@1
   320
default that column is assumed to be non-negative (in case of LP class)
alpar@1
   321
or binary (in case of MIP class).
alpar@1
   322
alpar@1
   323
\paragraph{Coefficient descriptors.} There must be exactly one
alpar@1
   324
coefficient descriptor line in the data file for each non-zero
alpar@1
   325
objective or constraint coefficient. This line has the following format:
alpar@1
   326
alpar@1
   327
\begin{verbatim}
alpar@1
   328
   a ROW COL VAL
alpar@1
   329
\end{verbatim}
alpar@1
   330
alpar@1
   331
The lower-case letter \verb|a| specifies that this is the coefficient
alpar@1
   332
descriptor line.
alpar@1
   333
alpar@1
   334
For objective coefficients the \verb|ROW| field must contain 0. For
alpar@1
   335
constraint coefficients the \verb|ROW| field specifies the row ordinal
alpar@1
   336
number, an integer between 1 and $m$, where $m$ is the number of rows
alpar@1
   337
in the problem instance.
alpar@1
   338
alpar@1
   339
The \verb|COL| field specifies the column ordinal number, an integer
alpar@1
   340
between 1 and $n$, where $n$ is the number of columns in the problem
alpar@1
   341
instance.
alpar@1
   342
alpar@1
   343
If both the \verb|ROW| and \verb|COL| fields contain 0, the line
alpar@1
   344
specifies the constant term (``shift'') of the objective function
alpar@1
   345
rather than objective coefficient.
alpar@1
   346
alpar@1
   347
The \verb|VAL| field contains a floating-point coefficient value (it is
alpar@1
   348
allowed to specify zero value in this field).
alpar@1
   349
alpar@1
   350
The number of constraint coefficient descriptor lines must be exactly
alpar@1
   351
the same as specified in the field \verb|NONZ| of the problem line.
alpar@1
   352
alpar@1
   353
\paragraph{Symbolic name descriptors.} There must be at most one
alpar@1
   354
symbolic name descriptor line for the problem instance, objective
alpar@1
   355
function, each row (constraint), and each column (variable). This line
alpar@1
   356
has one of the following formats:
alpar@1
   357
alpar@1
   358
\begin{verbatim}
alpar@1
   359
   n p NAME
alpar@1
   360
   n z NAME
alpar@1
   361
   n i ROW NAME
alpar@1
   362
   n j COL NAME
alpar@1
   363
\end{verbatim}
alpar@1
   364
alpar@1
   365
The lower-case letter \verb|n| specifies that this is the symbolic name
alpar@1
   366
descriptor line.
alpar@1
   367
alpar@1
   368
The next lower-case letter specifies which object should be assigned a
alpar@1
   369
symbolic name:
alpar@1
   370
alpar@1
   371
\verb|p| --- problem instance;
alpar@1
   372
alpar@1
   373
\verb|z| --- objective function;
alpar@1
   374
alpar@1
   375
\verb|i| --- row (constraint);
alpar@1
   376
alpar@1
   377
\verb|j| --- column (variable).
alpar@1
   378
alpar@1
   379
The \verb|ROW| field specifies the row ordinal number, an integer
alpar@1
   380
between 1 and $m$, where $m$ is the number of rows in the problem
alpar@1
   381
instance.
alpar@1
   382
alpar@1
   383
The \verb|COL| field specifies the column ordinal number, an integer
alpar@1
   384
between 1 and $n$, where $n$ is the number of columns in the problem
alpar@1
   385
instance.
alpar@1
   386
alpar@1
   387
The \verb|NAME| field contains the symbolic name, a sequence from 1 to
alpar@1
   388
255 arbitrary graphic ASCII characters, assigned to corresponding
alpar@1
   389
object.
alpar@1
   390
alpar@1
   391
\paragraph{End line.} There must be exactly one end line in the data
alpar@1
   392
file. This line must appear last in the file and has the following
alpar@1
   393
format:
alpar@1
   394
alpar@1
   395
\begin{verbatim}
alpar@1
   396
   e
alpar@1
   397
\end{verbatim}
alpar@1
   398
alpar@1
   399
The lower-case letter \verb|e| specifies that this is the end line.
alpar@1
   400
Anything that follows the end line is ignored by GLPK routines.
alpar@1
   401
alpar@1
   402
\subsubsection*{Example of data file in GLPK LP/MIP format}
alpar@1
   403
alpar@1
   404
The following example of a data file in GLPK LP/MIP format specifies
alpar@1
   405
the same LP problem as in Subsection ``Example of MPS file''.
alpar@1
   406
alpar@1
   407
\begin{center}
alpar@1
   408
\footnotesize\tt
alpar@1
   409
\begin{tabular}{l@{\hspace*{50pt}}}
alpar@1
   410
p lp min 8 7 48   \\
alpar@1
   411
n p PLAN          \\
alpar@1
   412
n z VALUE         \\
alpar@1
   413
i 1 f             \\
alpar@1
   414
n i 1 VALUE       \\
alpar@1
   415
i 2 s 2000        \\
alpar@1
   416
n i 2 YIELD       \\
alpar@1
   417
i 3 u 60          \\
alpar@1
   418
n i 3 FE          \\
alpar@1
   419
i 4 u 100         \\
alpar@1
   420
n i 4 CU          \\
alpar@1
   421
i 5 u 40          \\
alpar@1
   422
n i 5 MN          \\
alpar@1
   423
i 6 u 30          \\
alpar@1
   424
n i 6 MG          \\
alpar@1
   425
i 7 l 1500        \\
alpar@1
   426
n i 7 AL          \\
alpar@1
   427
i 8 d 250 300     \\
alpar@1
   428
n i 8 SI          \\
alpar@1
   429
j 1 d 0 200       \\
alpar@1
   430
n j 1 BIN1        \\
alpar@1
   431
j 2 d 0 2500      \\
alpar@1
   432
n j 2 BIN2        \\
alpar@1
   433
j 3 d 400 800     \\
alpar@1
   434
n j 3 BIN3        \\
alpar@1
   435
j 4 d 100 700     \\
alpar@1
   436
n j 4 BIN4        \\
alpar@1
   437
j 5 d 0 1500      \\
alpar@1
   438
n j 5 BIN5        \\
alpar@1
   439
n j 6 ALUM        \\
alpar@1
   440
n j 7 SILICON     \\
alpar@1
   441
a 0 1 0.03        \\
alpar@1
   442
a 0 2 0.08        \\
alpar@1
   443
a 0 3 0.17        \\
alpar@1
   444
a 0 4 0.12        \\
alpar@1
   445
a 0 5 0.15        \\
alpar@1
   446
a 0 6 0.21        \\
alpar@1
   447
a 0 7 0.38        \\
alpar@1
   448
a 1 1 0.03        \\
alpar@1
   449
a 1 2 0.08        \\
alpar@1
   450
a 1 3 0.17        \\
alpar@1
   451
a 1 4 0.12        \\
alpar@1
   452
a 1 5 0.15        \\
alpar@1
   453
a 1 6 0.21        \\
alpar@1
   454
\end{tabular}
alpar@1
   455
\begin{tabular}{|@{\hspace*{80pt}}l}
alpar@1
   456
a 1 7 0.38        \\
alpar@1
   457
a 2 1 1           \\
alpar@1
   458
a 2 2 1           \\
alpar@1
   459
a 2 3 1           \\
alpar@1
   460
a 2 4 1           \\
alpar@1
   461
a 2 5 1           \\
alpar@1
   462
a 2 6 1           \\
alpar@1
   463
a 2 7 1           \\
alpar@1
   464
a 3 1 0.15        \\
alpar@1
   465
a 3 2 0.04        \\
alpar@1
   466
a 3 3 0.02        \\
alpar@1
   467
a 3 4 0.04        \\
alpar@1
   468
a 3 5 0.02        \\
alpar@1
   469
a 3 6 0.01        \\
alpar@1
   470
a 3 7 0.03        \\
alpar@1
   471
a 4 1 0.03        \\
alpar@1
   472
a 4 2 0.05        \\
alpar@1
   473
a 4 3 0.08        \\
alpar@1
   474
a 4 4 0.02        \\
alpar@1
   475
a 4 5 0.06        \\
alpar@1
   476
a 4 6 0.01        \\
alpar@1
   477
a 5 1 0.02        \\
alpar@1
   478
a 5 2 0.04        \\
alpar@1
   479
a 5 3 0.01        \\
alpar@1
   480
a 5 4 0.02        \\
alpar@1
   481
a 5 5 0.02        \\
alpar@1
   482
a 6 1 0.02        \\
alpar@1
   483
a 6 2 0.03        \\
alpar@1
   484
a 6 5 0.01        \\
alpar@1
   485
a 7 1 0.7         \\
alpar@1
   486
a 7 2 0.75        \\
alpar@1
   487
a 7 3 0.8         \\
alpar@1
   488
a 7 4 0.75        \\
alpar@1
   489
a 7 5 0.8         \\
alpar@1
   490
a 7 6 0.97        \\
alpar@1
   491
a 8 1 0.02        \\
alpar@1
   492
a 8 2 0.06        \\
alpar@1
   493
a 8 3 0.08        \\
alpar@1
   494
a 8 4 0.12        \\
alpar@1
   495
a 8 5 0.02        \\
alpar@1
   496
a 8 6 0.01        \\
alpar@1
   497
a 8 7 0.97        \\
alpar@1
   498
e o f             \\
alpar@1
   499
\\
alpar@1
   500
\end{tabular}
alpar@1
   501
\end{center}
alpar@1
   502
alpar@1
   503
\newpage
alpar@1
   504
alpar@1
   505
\subsection{glp\_write\_prob---write problem data in GLPK format}
alpar@1
   506
alpar@1
   507
\subsubsection*{Synopsis}
alpar@1
   508
alpar@1
   509
\begin{verbatim}
alpar@1
   510
int glp_write_prob(glp_prob *P, int flags, const char *fname);
alpar@1
   511
\end{verbatim}
alpar@1
   512
alpar@1
   513
\subsubsection*{Description}
alpar@1
   514
alpar@1
   515
The routine \verb|glp_write_prob| writes problem data in the GLPK
alpar@1
   516
LP/MIP format to a text file. (For description of the GLPK LP/MIP
alpar@1
   517
format see Subsection ``Read problem data in GLPK format''.)
alpar@1
   518
alpar@1
   519
The parameter \verb|flags| is reserved for use in the future and should
alpar@1
   520
be specified as zero.
alpar@1
   521
alpar@1
   522
The character string \verb|fname| specifies a name of the text file to
alpar@1
   523
be written out. (If the file name ends with suffix `\verb|.gz|', the
alpar@1
   524
file is assumed to be compressed, in which case the routine
alpar@1
   525
\verb|glp_write_prob| performs automatic compression on writing it.)
alpar@1
   526
alpar@1
   527
\subsubsection*{Returns}
alpar@1
   528
alpar@1
   529
If the operation was successful, the routine \verb|glp_read_prob|
alpar@1
   530
returns zero. Otherwise, it prints an error message and returns
alpar@1
   531
non-zero.
alpar@1
   532
alpar@1
   533
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
alpar@1
   534
alpar@1
   535
\newpage
alpar@1
   536
alpar@1
   537
\section{Routines for processing MathProg models}
alpar@1
   538
alpar@1
   539
\subsection{Introduction}
alpar@1
   540
alpar@1
   541
GLPK supports the {\it GNU MathProg modeling language}.\footnote{The
alpar@1
   542
GNU MathProg modeling language is a subset of the AMPL language. For
alpar@1
   543
its detailed description see the document ``Modeling Language GNU
alpar@1
   544
MathProg: Language Reference'' included in the GLPK distribution.}
alpar@1
   545
As a rule, models written in MathProg are solved with the GLPK LP/MIP
alpar@1
   546
stand-alone solver \verb|glpsol| (see Appendix D) and do not need any
alpar@1
   547
programming with API routines. However, for various reasons the user
alpar@1
   548
may need to process MathProg models directly in his/her application
alpar@1
   549
program, in which case he/she may use API routines described in this
alpar@1
   550
section. These routines provide an interface to the {\it MathProg
alpar@1
   551
translator}, a component of GLPK, which translates MathProg models into
alpar@1
   552
an internal code and then interprets (executes) this code.
alpar@1
   553
alpar@1
   554
The processing of a model written in GNU MathProg includes several
alpar@1
   555
steps, which should be performed in the following order:
alpar@1
   556
alpar@1
   557
\begin{enumerate}
alpar@1
   558
\item{\it Allocating the workspace.}
alpar@1
   559
The translator allocates the workspace, an internal data structure used
alpar@1
   560
on all subsequent steps.
alpar@1
   561
\item{\it Reading model section.} The translator reads model section
alpar@1
   562
and, optionally, data section from a specified text file and translates
alpar@1
   563
them into the internal code. If necessary, on this step data section
alpar@1
   564
may be ignored.
alpar@1
   565
\item{\it Reading data section(s).} The translator reads one or more
alpar@1
   566
data sections from specified text file(s) and translates them into the
alpar@1
   567
internal code.
alpar@1
   568
\item{\it Generating the model.} The translator executes the internal
alpar@1
   569
code to evaluate the content of the model objects such as sets,
alpar@1
   570
parameters, variables, constraints, and objectives. On this step the
alpar@1
   571
execution is suspended at the solve statement.
alpar@1
   572
\item {\it Building the problem object.} The translator obtains all
alpar@1
   573
necessary information from the workspace and builds the standard
alpar@1
   574
problem object (that is, the program object of type \verb|glp_prob|).
alpar@1
   575
\item{\it Solving the problem.} On this step the problem object built
alpar@1
   576
on the previous step is passed to a solver, which solves the problem
alpar@1
   577
instance and stores its solution back to the problem object.
alpar@1
   578
\item{\it Postsolving the model.} The translator copies the solution
alpar@1
   579
from the problem object to the workspace and then executes the internal
alpar@1
   580
code from the solve statement to the end of the model. (If model has
alpar@1
   581
no solve statement, the translator does nothing on this step.)
alpar@1
   582
\item{\it Freeing the workspace.} The translator frees all the memory
alpar@1
   583
allocated to the workspace.
alpar@1
   584
\end{enumerate}
alpar@1
   585
alpar@1
   586
Note that the MathProg translator performs no error correction, so if
alpar@1
   587
any of steps 2 to 7 fails (due to errors in the model), the application
alpar@1
   588
program should terminate processing and go to step 8.
alpar@1
   589
alpar@1
   590
\subsubsection*{Example 1}
alpar@1
   591
alpar@1
   592
In this example the program reads model and data sections from input
alpar@1
   593
file \verb|egypt.mod|\footnote{This is an example model included in
alpar@1
   594
the GLPK distribution.} and writes the model to output file
alpar@1
   595
\verb|egypt.mps| in free MPS format (see Appendix B). No solution is
alpar@1
   596
performed.
alpar@1
   597
alpar@1
   598
\begin{small}
alpar@1
   599
\begin{verbatim}
alpar@1
   600
/* mplsamp1.c */
alpar@1
   601
alpar@1
   602
#include <stdio.h>
alpar@1
   603
#include <stdlib.h>
alpar@1
   604
#include <glpk.h>
alpar@1
   605
alpar@1
   606
int main(void)
alpar@1
   607
{     glp_prob *lp;
alpar@1
   608
      glp_tran *tran;
alpar@1
   609
      int ret;
alpar@1
   610
      lp = glp_create_prob();
alpar@1
   611
      tran = glp_mpl_alloc_wksp();
alpar@1
   612
      ret = glp_mpl_read_model(tran, "egypt.mod", 0);
alpar@1
   613
      if (ret != 0)
alpar@1
   614
      {  fprintf(stderr, "Error on translating model\n");
alpar@1
   615
         goto skip;
alpar@1
   616
      }
alpar@1
   617
      ret = glp_mpl_generate(tran, NULL);
alpar@1
   618
      if (ret != 0)
alpar@1
   619
      {  fprintf(stderr, "Error on generating model\n");
alpar@1
   620
         goto skip;
alpar@1
   621
      }
alpar@1
   622
      glp_mpl_build_prob(tran, lp);
alpar@1
   623
      ret = glp_write_mps(lp, GLP_MPS_FILE, NULL, "egypt.mps");
alpar@1
   624
      if (ret != 0)
alpar@1
   625
         fprintf(stderr, "Error on writing MPS file\n");
alpar@1
   626
skip: glp_mpl_free_wksp(tran);
alpar@1
   627
      glp_delete_prob(lp);
alpar@1
   628
      return 0;
alpar@1
   629
}
alpar@1
   630
alpar@1
   631
/* eof */
alpar@1
   632
\end{verbatim}
alpar@1
   633
\end{small}
alpar@1
   634
alpar@1
   635
\subsubsection*{Example 2}
alpar@1
   636
alpar@1
   637
In this example the program reads model section from file
alpar@1
   638
\verb|sudoku.mod|\footnote{This is an example model which is included
alpar@1
   639
in the GLPK distribution along with alternative data file
alpar@1
   640
{\tt sudoku.dat}.} ignoring data section in this file, reads alternative
alpar@1
   641
data section from file \verb|sudoku.dat|, solves the problem instance
alpar@1
   642
and passes the solution found back to the model.
alpar@1
   643
alpar@1
   644
\begin{small}
alpar@1
   645
\begin{verbatim}
alpar@1
   646
/* mplsamp2.c */
alpar@1
   647
alpar@1
   648
#include <stdio.h>
alpar@1
   649
#include <stdlib.h>
alpar@1
   650
#include <glpk.h>
alpar@1
   651
alpar@1
   652
int main(void)
alpar@1
   653
{     glp_prob *mip;
alpar@1
   654
      glp_tran *tran;
alpar@1
   655
      int ret;
alpar@1
   656
      mip = glp_create_prob();
alpar@1
   657
      tran = glp_mpl_alloc_wksp();
alpar@1
   658
      ret = glp_mpl_read_model(tran, "sudoku.mod", 1);
alpar@1
   659
      if (ret != 0)
alpar@1
   660
      {  fprintf(stderr, "Error on translating model\n");
alpar@1
   661
         goto skip;
alpar@1
   662
      }
alpar@1
   663
      ret = glp_mpl_read_data(tran, "sudoku.dat");
alpar@1
   664
      if (ret != 0)
alpar@1
   665
      {  fprintf(stderr, "Error on translating data\n");
alpar@1
   666
         goto skip;
alpar@1
   667
      }
alpar@1
   668
      ret = glp_mpl_generate(tran, NULL);
alpar@1
   669
      if (ret != 0)
alpar@1
   670
      {  fprintf(stderr, "Error on generating model\n");
alpar@1
   671
         goto skip;
alpar@1
   672
      }
alpar@1
   673
      glp_mpl_build_prob(tran, mip);
alpar@1
   674
      glp_simplex(mip, NULL);
alpar@1
   675
      glp_intopt(mip, NULL);
alpar@1
   676
      ret = glp_mpl_postsolve(tran, mip, GLP_MIP);
alpar@1
   677
      if (ret != 0)
alpar@1
   678
         fprintf(stderr, "Error on postsolving model\n");
alpar@1
   679
skip: glp_mpl_free_wksp(tran);
alpar@1
   680
      glp_delete_prob(mip);
alpar@1
   681
      return 0;
alpar@1
   682
}
alpar@1
   683
alpar@1
   684
/* eof */
alpar@1
   685
\end{verbatim}
alpar@1
   686
\end{small}
alpar@1
   687
alpar@1
   688
\subsection{glp\_mpl\_alloc\_wksp---allocate the translator workspace}
alpar@1
   689
alpar@1
   690
\subsubsection*{Synopsis}
alpar@1
   691
alpar@1
   692
\begin{verbatim}
alpar@1
   693
glp_tran *glp_mpl_alloc_wksp(void);
alpar@1
   694
\end{verbatim}
alpar@1
   695
alpar@1
   696
\subsubsection*{Description}
alpar@1
   697
alpar@1
   698
The routine \verb|glp_mpl_alloc_wksp| allocates the MathProg translator
alpar@1
   699
work\-space. (Note that multiple instances of the workspace may be
alpar@1
   700
allocated, if necessary.)
alpar@1
   701
alpar@1
   702
\subsubsection*{Returns}
alpar@1
   703
alpar@1
   704
The routine returns a pointer to the workspace, which should be used in
alpar@1
   705
all subsequent operations.
alpar@1
   706
alpar@1
   707
\subsection{glp\_mpl\_read\_model---read and translate model section}
alpar@1
   708
alpar@1
   709
\subsubsection*{Synopsis}
alpar@1
   710
alpar@1
   711
\begin{verbatim}
alpar@1
   712
int glp_mpl_read_model(glp_tran *tran, const char *fname,
alpar@1
   713
      int skip);
alpar@1
   714
\end{verbatim}
alpar@1
   715
alpar@1
   716
\subsubsection*{Description}
alpar@1
   717
alpar@1
   718
The routine \verb|glp_mpl_read_model| reads model section and,
alpar@1
   719
optionally, data section, which may follow the model section, from a
alpar@1
   720
text file, whose name is the character string \verb|fname|, performs
alpar@1
   721
translation of model statements and data blocks, and stores all the
alpar@1
   722
information in the workspace.
alpar@1
   723
alpar@1
   724
The parameter \verb|skip| is a flag. If the input file contains the
alpar@1
   725
data section and this flag is non-zero, the data section is not read as
alpar@1
   726
if there were no data section and a warning message is printed. This
alpar@1
   727
allows reading data section(s) from other file(s).
alpar@1
   728
alpar@1
   729
\subsubsection*{Returns}
alpar@1
   730
alpar@1
   731
If the operation is successful, the routine returns zero. Otherwise
alpar@1
   732
the routine prints an error message and returns non-zero.
alpar@1
   733
alpar@1
   734
\subsection{glp\_mpl\_read\_data---read and translate data section}
alpar@1
   735
alpar@1
   736
\subsubsection*{Synopsis}
alpar@1
   737
alpar@1
   738
\begin{verbatim}
alpar@1
   739
int glp_mpl_read_data(glp_tran *tran, const char *fname);
alpar@1
   740
\end{verbatim}
alpar@1
   741
alpar@1
   742
\subsubsection*{Description}
alpar@1
   743
alpar@1
   744
The routine \verb|glp_mpl_read_data| reads data section from a text
alpar@1
   745
file, whose name is the character string \verb|fname|, performs
alpar@1
   746
translation of data blocks, and stores the data read in the translator
alpar@1
   747
workspace. If necessary, this routine may be called more than once.
alpar@1
   748
alpar@1
   749
\subsubsection*{Returns}
alpar@1
   750
alpar@1
   751
If the operation is successful, the routine returns zero. Otherwise
alpar@1
   752
the routine prints an error message and returns non-zero.
alpar@1
   753
alpar@1
   754
\subsection{glp\_mpl\_generate---generate the model}
alpar@1
   755
alpar@1
   756
\subsubsection*{Synopsis}
alpar@1
   757
alpar@1
   758
\begin{verbatim}
alpar@1
   759
int glp_mpl_generate(glp_tran *tran, const char *fname);
alpar@1
   760
\end{verbatim}
alpar@1
   761
alpar@1
   762
\subsubsection*{Description}
alpar@1
   763
alpar@1
   764
The routine \verb|glp_mpl_generate| generates the model using its
alpar@1
   765
description stored in the translator workspace. This operation means
alpar@1
   766
generating all variables, constraints, and objectives, executing check
alpar@1
   767
and display statements, which precede the solve statement (if it is
alpar@1
   768
presented).
alpar@1
   769
alpar@1
   770
The character string \verb|fname| specifies the name of an output text
alpar@1
   771
file, to which output produced by display statements should be written.
alpar@1
   772
If \verb|fname| is \verb|NULL|, the output is sent to the terminal.
alpar@1
   773
alpar@1
   774
\subsubsection*{Returns}
alpar@1
   775
alpar@1
   776
If the operation is successful, the routine returns zero. Otherwise
alpar@1
   777
the routine prints an error message and returns non-zero.
alpar@1
   778
alpar@1
   779
\subsection{glp\_mpl\_build\_prob---build problem instance from the
alpar@1
   780
model}
alpar@1
   781
alpar@1
   782
\subsubsection*{Synopsis}
alpar@1
   783
alpar@1
   784
\begin{verbatim}
alpar@1
   785
void glp_mpl_build_prob(glp_tran *tran, glp_prob *prob);
alpar@1
   786
\end{verbatim}
alpar@1
   787
alpar@1
   788
\subsubsection*{Description}
alpar@1
   789
alpar@1
   790
The routine \verb|glp_mpl_build_prob| obtains all necessary information
alpar@1
   791
from the translator workspace and stores it in the specified problem
alpar@1
   792
object \verb|prob|. Note that before building the current content of
alpar@1
   793
the problem object is erased with the routine \verb|glp_erase_prob|.
alpar@1
   794
alpar@1
   795
\subsection{glp\_mpl\_postsolve---postsolve the model}
alpar@1
   796
alpar@1
   797
\subsubsection*{Synopsis}
alpar@1
   798
alpar@1
   799
\begin{verbatim}
alpar@1
   800
int glp_mpl_postsolve(glp_tran *tran, glp_prob *prob,
alpar@1
   801
      int sol);
alpar@1
   802
\end{verbatim}
alpar@1
   803
alpar@1
   804
\subsubsection*{Description}
alpar@1
   805
alpar@1
   806
The routine \verb|glp_mpl_postsolve| copies the solution from the
alpar@1
   807
specified problem object \verb|prob| to the translator workspace and
alpar@1
   808
then executes all the remaining model statements, which follow the
alpar@1
   809
solve statement.
alpar@1
   810
alpar@1
   811
The parameter \verb|sol| specifies which solution should be copied
alpar@1
   812
from the problem object to the workspace as follows:
alpar@1
   813
alpar@1
   814
\begin{tabular}{@{}ll}
alpar@1
   815
\verb|GLP_SOL| & basic solution; \\
alpar@1
   816
\verb|GLP_IPT| & interior-point solution; \\
alpar@1
   817
\verb|GLP_MIP| & mixed integer solution. \\
alpar@1
   818
\end{tabular}
alpar@1
   819
alpar@1
   820
\subsubsection*{Returns}
alpar@1
   821
alpar@1
   822
If the operation is successful, the routine returns zero. Otherwise
alpar@1
   823
the routine prints an error message and returns non-zero.
alpar@1
   824
alpar@1
   825
\subsection{glp\_mpl\_free\_wksp---free the translator workspace}
alpar@1
   826
alpar@1
   827
\subsubsection*{Synopsis}
alpar@1
   828
alpar@1
   829
\begin{verbatim}
alpar@1
   830
void glp_mpl_free_wksp(glp_tran *tran);
alpar@1
   831
\end{verbatim}
alpar@1
   832
alpar@1
   833
\subsubsection*{Description}
alpar@1
   834
alpar@1
   835
The routine \verb|glp_mpl_free_wksp| frees all the memory allocated to
alpar@1
   836
the translator workspace. It also frees all other resources, which are
alpar@1
   837
still used by the translator.
alpar@1
   838
alpar@1
   839
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
alpar@1
   840
alpar@1
   841
\newpage
alpar@1
   842
alpar@1
   843
\section{Problem solution reading/writing routines}
alpar@1
   844
alpar@1
   845
\subsection{glp\_print\_sol---write basic solution in printable format}
alpar@1
   846
alpar@1
   847
\subsubsection*{Synopsis}
alpar@1
   848
alpar@1
   849
\begin{verbatim}
alpar@1
   850
int glp_print_sol(glp_prob *lp, const char *fname);
alpar@1
   851
\end{verbatim}
alpar@1
   852
alpar@1
   853
\subsubsection*{Description}
alpar@1
   854
alpar@1
   855
The routine \verb|glp_print_sol writes| the current basic solution of
alpar@1
   856
an LP problem, which is specified by the pointer \verb|lp|, to a text
alpar@1
   857
file, whose name is the character string \verb|fname|, in printable
alpar@1
   858
format.
alpar@1
   859
alpar@1
   860
Information reported by the routine \verb|glp_print_sol| is intended
alpar@1
   861
mainly for visual analysis.
alpar@1
   862
alpar@1
   863
\subsubsection*{Returns}
alpar@1
   864
alpar@1
   865
If no errors occurred, the routine returns zero. Otherwise the routine
alpar@1
   866
prints an error message and returns non-zero.
alpar@1
   867
alpar@1
   868
\subsection{glp\_read\_sol---read basic solution from text file}
alpar@1
   869
alpar@1
   870
\subsubsection*{Synopsis}
alpar@1
   871
alpar@1
   872
\begin{verbatim}
alpar@1
   873
int glp_read_sol(glp_prob *lp, const char *fname);
alpar@1
   874
\end{verbatim}
alpar@1
   875
alpar@1
   876
\subsubsection*{Description}
alpar@1
   877
alpar@1
   878
The routine \verb|glp_read_sol| reads basic solution from a text file
alpar@1
   879
whose name is specified by the parameter \verb|fname| into the problem
alpar@1
   880
object.
alpar@1
   881
alpar@1
   882
For the file format see description of the routine \verb|glp_write_sol|.
alpar@1
   883
alpar@1
   884
\subsubsection*{Returns}
alpar@1
   885
alpar@1
   886
On success the routine returns zero, otherwise non-zero.
alpar@1
   887
alpar@1
   888
\newpage
alpar@1
   889
alpar@1
   890
\subsection{glp\_write\_sol---write basic solution to text file}
alpar@1
   891
alpar@1
   892
\subsubsection*{Synopsis}
alpar@1
   893
alpar@1
   894
\begin{verbatim}
alpar@1
   895
int glp_write_sol(glp_prob *lp, const char *fname);
alpar@1
   896
\end{verbatim}
alpar@1
   897
alpar@1
   898
\subsubsection*{Description}
alpar@1
   899
alpar@1
   900
The routine \verb|glp_write_sol| writes the current basic solution to a
alpar@1
   901
text file whose name is specified by the parameter \verb|fname|. This
alpar@1
   902
file can be read back with the routine \verb|glp_read_sol|.
alpar@1
   903
alpar@1
   904
\subsubsection*{Returns}
alpar@1
   905
alpar@1
   906
On success the routine returns zero, otherwise non-zero.
alpar@1
   907
alpar@1
   908
\subsubsection*{File format}
alpar@1
   909
alpar@1
   910
The file created by the routine \verb|glp_write_sol| is a plain text
alpar@1
   911
file, which contains the following information:
alpar@1
   912
alpar@1
   913
\begin{verbatim}
alpar@1
   914
   m n
alpar@1
   915
   p_stat d_stat obj_val
alpar@1
   916
   r_stat[1] r_prim[1] r_dual[1]
alpar@1
   917
   . . .
alpar@1
   918
   r_stat[m] r_prim[m] r_dual[m]
alpar@1
   919
   c_stat[1] c_prim[1] c_dual[1]
alpar@1
   920
   . . .
alpar@1
   921
   c_stat[n] c_prim[n] c_dual[n]
alpar@1
   922
\end{verbatim}
alpar@1
   923
alpar@1
   924
\noindent
alpar@1
   925
where:
alpar@1
   926
alpar@1
   927
\noindent
alpar@1
   928
$m$ is the number of rows (auxiliary variables);
alpar@1
   929
alpar@1
   930
\noindent
alpar@1
   931
$n$ is the number of columns (structural variables);
alpar@1
   932
alpar@1
   933
\noindent
alpar@1
   934
\verb|p_stat| is the primal status of the basic solution
alpar@1
   935
(\verb|GLP_UNDEF| = 1, \verb|GLP_FEAS| = 2, \verb|GLP_INFEAS| = 3, or
alpar@1
   936
\verb|GLP_NOFEAS| = 4);
alpar@1
   937
alpar@1
   938
\noindent
alpar@1
   939
\verb|d_stat| is the dual status of the basic solution
alpar@1
   940
(\verb|GLP_UNDEF| = 1, \verb|GLP_FEAS| = 2, \verb|GLP_INFEAS| = 3, or
alpar@1
   941
\verb|GLP_NOFEAS| = 4);
alpar@1
   942
alpar@1
   943
\noindent
alpar@1
   944
\verb|obj_val| is the objective value;
alpar@1
   945
alpar@1
   946
\noindent
alpar@1
   947
\verb|r_stat[i]|, $i=1,\dots,m$, is the status of $i$-th row
alpar@1
   948
(\verb|GLP_BS| = 1, \verb|GLP_NL| = 2, \verb|GLP_NU| = 3,
alpar@1
   949
\verb|GLP_NF| = 4, or \verb|GLP_NS| = 5);
alpar@1
   950
alpar@1
   951
\noindent
alpar@1
   952
\verb|r_prim[i]|, $i=1,\dots,m$, is the primal value of $i$-th row;
alpar@1
   953
alpar@1
   954
\noindent
alpar@1
   955
\verb|r_dual[i]|, $i=1,\dots,m$, is the dual value of $i$-th row;
alpar@1
   956
alpar@1
   957
\noindent
alpar@1
   958
\verb|c_stat[j]|, $j=1,\dots,n$, is the status of $j$-th column
alpar@1
   959
(\verb|GLP_BS| = 1, \verb|GLP_NL| = 2, \verb|GLP_NU| = 3,
alpar@1
   960
\verb|GLP_NF| = 4, or \verb|GLP_NS| = 5);
alpar@1
   961
alpar@1
   962
\noindent
alpar@1
   963
\verb|c_prim[j]|, $j=1,\dots,n$, is the primal value of $j$-th column;
alpar@1
   964
alpar@1
   965
\noindent
alpar@1
   966
\verb|c_dual[j]|, $j=1,\dots,n$, is the dual value of $j$-th column.
alpar@1
   967
alpar@1
   968
\subsection{glp\_print\_ipt---write interior-point solution in
alpar@1
   969
printable format}
alpar@1
   970
alpar@1
   971
\subsubsection*{Synopsis}
alpar@1
   972
alpar@1
   973
\begin{verbatim}
alpar@1
   974
int glp_print_ipt(glp_prob *lp, const char *fname);
alpar@1
   975
\end{verbatim}
alpar@1
   976
alpar@1
   977
\subsubsection*{Description}
alpar@1
   978
alpar@1
   979
The routine \verb|glp_print_ipt| writes the current interior point
alpar@1
   980
solution  of an LP problem, which the parameter \verb|lp| points to, to
alpar@1
   981
a text file, whose name is the character string \verb|fname|, in
alpar@1
   982
printable format.
alpar@1
   983
alpar@1
   984
Information reported by the routine \verb|glp_print_ipt| is intended
alpar@1
   985
mainly for visual analysis.
alpar@1
   986
alpar@1
   987
\subsubsection*{Returns}
alpar@1
   988
alpar@1
   989
If no errors occurred, the routine returns zero. Otherwise the routine
alpar@1
   990
prints an error message and returns non-zero.
alpar@1
   991
alpar@1
   992
\subsection{glp\_read\_ipt---read interior-point solution from text
alpar@1
   993
file}
alpar@1
   994
alpar@1
   995
\subsubsection*{Synopsis}
alpar@1
   996
alpar@1
   997
\begin{verbatim}
alpar@1
   998
int glp_read_ipt(glp_prob *lp, const char *fname);
alpar@1
   999
\end{verbatim}
alpar@1
  1000
alpar@1
  1001
\subsubsection*{Description}
alpar@1
  1002
alpar@1
  1003
The routine \verb|glp_read_ipt| reads interior-point solution from a
alpar@1
  1004
text file whose name is specified by the parameter \verb|fname| into the
alpar@1
  1005
problem object.
alpar@1
  1006
alpar@1
  1007
For the file format see description of the routine \verb|glp_write_ipt|.
alpar@1
  1008
alpar@1
  1009
\subsubsection*{Returns}
alpar@1
  1010
alpar@1
  1011
On success the routine returns zero, otherwise non-zero.
alpar@1
  1012
alpar@1
  1013
\subsection{glp\_write\_ipt---write interior-point solution to text
alpar@1
  1014
file}
alpar@1
  1015
alpar@1
  1016
\subsubsection*{Synopsis}
alpar@1
  1017
alpar@1
  1018
\begin{verbatim}
alpar@1
  1019
int glp_write_ipt(glp_prob *lp, const char *fname);
alpar@1
  1020
\end{verbatim}
alpar@1
  1021
alpar@1
  1022
\subsubsection*{Description}
alpar@1
  1023
alpar@1
  1024
The routine \verb|glp_write_ipt| writes the current interior-point
alpar@1
  1025
solution to a text file whose name is specified by the parameter
alpar@1
  1026
\verb|fname|. This file can be read back with the routine
alpar@1
  1027
\verb|glp_read_ipt|.
alpar@1
  1028
alpar@1
  1029
\subsubsection*{Returns}
alpar@1
  1030
alpar@1
  1031
On success the routine returns zero, otherwise non-zero.
alpar@1
  1032
alpar@1
  1033
\subsubsection*{File format}
alpar@1
  1034
alpar@1
  1035
The file created by the routine \verb|glp_write_ipt| is a plain text
alpar@1
  1036
file, which contains the following information:
alpar@1
  1037
alpar@1
  1038
\begin{verbatim}
alpar@1
  1039
   m n
alpar@1
  1040
   stat obj_val
alpar@1
  1041
   r_prim[1] r_dual[1]
alpar@1
  1042
   . . .
alpar@1
  1043
   r_prim[m] r_dual[m]
alpar@1
  1044
   c_prim[1] c_dual[1]
alpar@1
  1045
   . . .
alpar@1
  1046
   c_prim[n] c_dual[n]
alpar@1
  1047
\end{verbatim}
alpar@1
  1048
alpar@1
  1049
\noindent
alpar@1
  1050
where:
alpar@1
  1051
alpar@1
  1052
\noindent
alpar@1
  1053
$m$ is the number of rows (auxiliary variables);
alpar@1
  1054
alpar@1
  1055
\noindent
alpar@1
  1056
$n$ is the number of columns (structural variables);
alpar@1
  1057
alpar@1
  1058
\noindent
alpar@1
  1059
\verb|stat| is the solution status (\verb|GLP_UNDEF| = 1 or
alpar@1
  1060
\verb|GLP_OPT| = 5);
alpar@1
  1061
alpar@1
  1062
\noindent
alpar@1
  1063
\verb|obj_val| is the objective value;
alpar@1
  1064
alpar@1
  1065
\noindent
alpar@1
  1066
\verb|r_prim[i]|, $i=1,\dots,m$, is the primal value of $i$-th row;
alpar@1
  1067
alpar@1
  1068
\noindent
alpar@1
  1069
\verb|r_dual[i]|, $i=1,\dots,m$, is the dual value of $i$-th row;
alpar@1
  1070
alpar@1
  1071
\noindent
alpar@1
  1072
\verb|c_prim[j]|, $j=1,\dots,n$, is the primal value of $j$-th column;
alpar@1
  1073
alpar@1
  1074
\noindent
alpar@1
  1075
\verb|c_dual[j]|, $j=1,\dots,n$, is the dual value of $j$-th column.
alpar@1
  1076
alpar@1
  1077
\subsection{glp\_print\_mip---write MIP solution in printable format}
alpar@1
  1078
alpar@1
  1079
\subsubsection*{Synopsis}
alpar@1
  1080
alpar@1
  1081
\begin{verbatim}
alpar@1
  1082
int glp_print_mip(glp_prob *lp, const char *fname);
alpar@1
  1083
\end{verbatim}
alpar@1
  1084
alpar@1
  1085
\subsubsection*{Description}
alpar@1
  1086
alpar@1
  1087
The routine \verb|glp_print_mip| writes a best known integer solution
alpar@1
  1088
of a MIP problem, which is specified by the pointer \verb|lp|, to a text
alpar@1
  1089
file, whose name is the character string \verb|fname|, in printable
alpar@1
  1090
format.
alpar@1
  1091
alpar@1
  1092
Information reported by the routine \verb|glp_print_mip| is intended
alpar@1
  1093
mainly for visual analysis.
alpar@1
  1094
alpar@1
  1095
\subsubsection*{Returns}
alpar@1
  1096
alpar@1
  1097
If no errors occurred, the routine returns zero. Otherwise the routine
alpar@1
  1098
prints an error message and returns non-zero.
alpar@1
  1099
alpar@1
  1100
\newpage
alpar@1
  1101
alpar@1
  1102
\subsection{glp\_read\_mip---read MIP solution from text file}
alpar@1
  1103
alpar@1
  1104
\subsubsection*{Synopsis}
alpar@1
  1105
alpar@1
  1106
\begin{verbatim}
alpar@1
  1107
int glp_read_mip(glp_prob *mip, const char *fname);
alpar@1
  1108
\end{verbatim}
alpar@1
  1109
alpar@1
  1110
\subsubsection*{Description}
alpar@1
  1111
alpar@1
  1112
The routine \verb|glp_read_mip| reads MIP solution from a text file
alpar@1
  1113
whose name is specified by the parameter \verb|fname| into the problem
alpar@1
  1114
object.
alpar@1
  1115
alpar@1
  1116
For the file format see description of the routine \verb|glp_write_mip|.
alpar@1
  1117
alpar@1
  1118
\subsubsection*{Returns}
alpar@1
  1119
alpar@1
  1120
On success the routine returns zero, otherwise non-zero.
alpar@1
  1121
alpar@1
  1122
\subsection{glp\_write\_mip---write MIP solution to text file}
alpar@1
  1123
alpar@1
  1124
\subsubsection*{Synopsis}
alpar@1
  1125
alpar@1
  1126
\begin{verbatim}
alpar@1
  1127
int glp_write_mip(glp_prob *mip, const char *fname);
alpar@1
  1128
\end{verbatim}
alpar@1
  1129
alpar@1
  1130
\subsubsection*{Description}
alpar@1
  1131
alpar@1
  1132
The routine \verb|glp_write_mip| writes the current MIP solution to a
alpar@1
  1133
text file whose name is specified by the parameter \verb|fname|. This
alpar@1
  1134
file can be read back with the routine \verb|glp_read_mip|.
alpar@1
  1135
alpar@1
  1136
\subsubsection*{Returns}
alpar@1
  1137
alpar@1
  1138
On success the routine returns zero, otherwise non-zero.
alpar@1
  1139
alpar@1
  1140
\subsubsection*{File format}
alpar@1
  1141
alpar@1
  1142
The file created by the routine \verb|glp_write_sol| is a plain text
alpar@1
  1143
file, which contains the following information:
alpar@1
  1144
alpar@1
  1145
\begin{verbatim}
alpar@1
  1146
   m n
alpar@1
  1147
   stat obj_val
alpar@1
  1148
   r_val[1]
alpar@1
  1149
   . . .
alpar@1
  1150
   r_val[m]
alpar@1
  1151
   c_val[1]
alpar@1
  1152
   . . .
alpar@1
  1153
   c_val[n]
alpar@1
  1154
\end{verbatim}
alpar@1
  1155
alpar@1
  1156
\noindent
alpar@1
  1157
where:
alpar@1
  1158
alpar@1
  1159
\noindent
alpar@1
  1160
$m$ is the number of rows (auxiliary variables);
alpar@1
  1161
alpar@1
  1162
\noindent
alpar@1
  1163
$n$ is the number of columns (structural variables);
alpar@1
  1164
alpar@1
  1165
\noindent
alpar@1
  1166
\verb|stat| is the solution status (\verb|GLP_UNDEF| = 1,
alpar@1
  1167
\verb|GLP_FEAS| = 2, \verb|GLP_NOFEAS| = 4, or \verb|GLP_OPT| = 5);
alpar@1
  1168
alpar@1
  1169
\noindent
alpar@1
  1170
\verb|obj_val| is the objective value;
alpar@1
  1171
alpar@1
  1172
\noindent
alpar@1
  1173
\verb|r_val[i]|, $i=1,\dots,m$, is the value of $i$-th row;
alpar@1
  1174
alpar@1
  1175
\noindent
alpar@1
  1176
\verb|c_val[j]|, $j=1,\dots,n$, is the value of $j$-th column.
alpar@1
  1177
alpar@1
  1178
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
alpar@1
  1179
alpar@1
  1180
\newpage
alpar@1
  1181
alpar@1
  1182
\section{Post-optimal analysis routines}
alpar@1
  1183
alpar@1
  1184
\subsection{glp\_print\_ranges---print sensitivity analysis report}
alpar@1
  1185
alpar@1
  1186
\subsubsection*{Synopsis}
alpar@1
  1187
alpar@1
  1188
\begin{verbatim}
alpar@1
  1189
int glp_print_ranges(glp_prob *P, int len, const int list[],
alpar@1
  1190
      int flags, const char *fname);
alpar@1
  1191
\end{verbatim}
alpar@1
  1192
alpar@1
  1193
\subsubsection*{Description}
alpar@1
  1194
alpar@1
  1195
The routine \verb|glp_print_ranges| performs sensitivity analysis of
alpar@1
  1196
current optimal basic solution and writes the analysis report in
alpar@1
  1197
human-readable format to a text file, whose name is the character
alpar@1
  1198
string {\it fname}. (Detailed description of the report structure is
alpar@1
  1199
given below.)
alpar@1
  1200
alpar@1
  1201
The parameter {\it len} specifies the length of the row/column list.
alpar@1
  1202
alpar@1
  1203
The array {\it list} specifies ordinal number of rows and columns to be
alpar@1
  1204
analyzed. The ordinal numbers should be passed in locations
alpar@1
  1205
{\it list}[1], {\it list}[2], \dots, {\it list}[{\it len}]. Ordinal
alpar@1
  1206
numbers from 1 to $m$ refer to rows, and ordinal numbers from $m+1$ to
alpar@1
  1207
$m+n$ refer to columns, where $m$ and $n$ are, resp., the total number
alpar@1
  1208
of rows and columns in the problem object. Rows and columns appear in
alpar@1
  1209
the analysis report in the same order as they follow in the array list.
alpar@1
  1210
alpar@1
  1211
It is allowed to specify $len=0$, in which case the array {\it list} is
alpar@1
  1212
not used (so it can be specified as \verb|NULL|), and the routine
alpar@1
  1213
performs analysis for all rows and columns of the problem object.
alpar@1
  1214
alpar@1
  1215
The parameter {\it flags} is reserved for use in the future and must be
alpar@1
  1216
specified as zero.
alpar@1
  1217
alpar@1
  1218
On entry to the routine \verb|glp_print_ranges| the current basic
alpar@1
  1219
solution must be optimal and the basis factorization must exist.
alpar@1
  1220
The application program can check that with the routine
alpar@1
  1221
\verb|glp_bf_exists|, and if the factorization does
alpar@1
  1222
not exist, compute it with the routine \verb|glp_factorize|. Note that
alpar@1
  1223
if the LP preprocessor is not used, on normal exit from the simplex
alpar@1
  1224
solver routine \verb|glp_simplex| the basis factorization always exists.
alpar@1
  1225
alpar@1
  1226
\subsubsection*{Returns}
alpar@1
  1227
alpar@1
  1228
If the operation was successful, the routine \verb|glp_print_ranges|
alpar@1
  1229
returns zero. Otherwise, it prints an error message and returns
alpar@1
  1230
non-zero.
alpar@1
  1231
alpar@1
  1232
\subsubsection*{Analysis report example}
alpar@1
  1233
alpar@1
  1234
An example of the sensitivity analysis report is shown on the next two
alpar@1
  1235
pages. This example corresponds to the example of LP problem described
alpar@1
  1236
in Subsection ``Example of MPS file''.
alpar@1
  1237
alpar@1
  1238
\subsubsection*{Structure of the analysis report}
alpar@1
  1239
alpar@1
  1240
For each row and column specified in the array {\it list} the routine
alpar@1
  1241
prints two lines containing generic information and analysis
alpar@1
  1242
information, which depends on the status of corresponding row or column.
alpar@1
  1243
alpar@1
  1244
Note that analysis of a row is analysis of its auxiliary variable,
alpar@1
  1245
which is equal to the row linear form $\sum a_jx_j$, and analysis of
alpar@1
  1246
a column is analysis of corresponding structural variable. Therefore,
alpar@1
  1247
formally, on performing the sensitivity analysis there is no difference
alpar@1
  1248
between rows and columns.
alpar@1
  1249
alpar@1
  1250
\bigskip
alpar@1
  1251
alpar@1
  1252
\noindent
alpar@1
  1253
{\it Generic information}
alpar@1
  1254
alpar@1
  1255
\medskip
alpar@1
  1256
alpar@1
  1257
\noindent
alpar@1
  1258
{\tt No.} is the row or column ordinal number in the problem object.
alpar@1
  1259
Rows are numbered from 1 to $m$, and columns are numbered from 1 to $n$,
alpar@1
  1260
where $m$ and $n$ are, resp., the total number of rows and columns in
alpar@1
  1261
the problem object.
alpar@1
  1262
alpar@1
  1263
\medskip
alpar@1
  1264
alpar@1
  1265
\noindent
alpar@1
  1266
{\tt Row name} is the symbolic name assigned to the row. If the row has
alpar@1
  1267
no name assigned, this field contains blanks.
alpar@1
  1268
alpar@1
  1269
\medskip
alpar@1
  1270
alpar@1
  1271
\noindent
alpar@1
  1272
{\tt Column name} is the symbolic name assigned to the column. If the
alpar@1
  1273
column has no name assigned, this field contains blanks.
alpar@1
  1274
alpar@1
  1275
\medskip
alpar@1
  1276
alpar@1
  1277
\noindent
alpar@1
  1278
{\tt St} is the status of the row or column in the optimal solution:
alpar@1
  1279
alpar@1
  1280
{\tt BS} --- non-active constraint (row), basic column;
alpar@1
  1281
alpar@1
  1282
{\tt NL} --- inequality constraint having its lower right-hand side
alpar@1
  1283
active (row), non-basic column having its lower bound active;
alpar@1
  1284
alpar@1
  1285
{\tt NU} --- inequality constraint having its upper right-hand side
alpar@1
  1286
active (row), non-basic column having its upper bound active;
alpar@1
  1287
alpar@1
  1288
{\tt NS} --- active equality constraint (row), non-basic fixed column.
alpar@1
  1289
alpar@1
  1290
{\tt NF} --- active free row, non-basic free (unbounded) column. (This
alpar@1
  1291
case means that the optimal solution is dual degenerate.)
alpar@1
  1292
alpar@1
  1293
\medskip
alpar@1
  1294
alpar@1
  1295
\noindent
alpar@1
  1296
{\tt Activity} is the (primal) value of the auxiliary variable (row) or
alpar@1
  1297
structural variable (column) in the optimal solution.
alpar@1
  1298
alpar@1
  1299
\medskip
alpar@1
  1300
alpar@1
  1301
\noindent
alpar@1
  1302
{\tt Slack} is the (primal) value of the row slack variable.
alpar@1
  1303
alpar@1
  1304
\medskip
alpar@1
  1305
alpar@1
  1306
\noindent
alpar@1
  1307
{\tt Obj coef} is the objective coefficient of the column (structural
alpar@1
  1308
variable).
alpar@1
  1309
alpar@1
  1310
\begin{landscape}
alpar@1
  1311
\begin{scriptsize}
alpar@1
  1312
\begin{verbatim}
alpar@1
  1313
GLPK 4.42 - SENSITIVITY ANALYSIS REPORT                                                                         Page   1
alpar@1
  1314
alpar@1
  1315
Problem:    PLAN
alpar@1
  1316
Objective:  VALUE = 296.2166065 (MINimum)
alpar@1
  1317
alpar@1
  1318
   No. Row name     St      Activity         Slack   Lower bound       Activity      Obj coef  Obj value at Limiting
alpar@1
  1319
                                          Marginal   Upper bound          range         range   break point variable
alpar@1
  1320
------ ------------ -- ------------- ------------- -------------  ------------- ------------- ------------- ------------
alpar@1
  1321
     1 VALUE        BS     296.21661    -296.21661          -Inf      299.25255      -1.00000        .      MN
alpar@1
  1322
                                            .               +Inf      296.21661          +Inf          +Inf
alpar@1
  1323
alpar@1
  1324
     2 YIELD        NS    2000.00000        .         2000.00000     1995.06864          -Inf     296.28365 BIN3
alpar@1
  1325
                                           -.01360    2000.00000     2014.03479          +Inf     296.02579 CU
alpar@1
  1326
alpar@1
  1327
     3 FE           NU      60.00000        .               -Inf       55.89016          -Inf     306.77162 BIN4
alpar@1
  1328
                                          -2.56823      60.00000       62.69978       2.56823     289.28294 BIN3
alpar@1
  1329
alpar@1
  1330
     4 CU           BS      83.96751      16.03249          -Inf       93.88467       -.30613     270.51157 MN
alpar@1
  1331
                                            .          100.00000       79.98213        .21474     314.24798 BIN5
alpar@1
  1332
alpar@1
  1333
     5 MN           NU      40.00000        .               -Inf       34.42336          -Inf     299.25255 BIN4
alpar@1
  1334
                                           -.54440      40.00000       41.68691        .54440     295.29825 BIN3
alpar@1
  1335
alpar@1
  1336
     6 MG           BS      19.96029      10.03971          -Inf       24.74427      -1.79618     260.36433 BIN1
alpar@1
  1337
                                            .           30.00000        9.40292        .28757     301.95652 MN
alpar@1
  1338
alpar@1
  1339
     7 AL           NL    1500.00000        .         1500.00000     1485.78425       -.25199     292.63444 CU
alpar@1
  1340
                                            .25199          +Inf     1504.92126          +Inf     297.45669 BIN3
alpar@1
  1341
alpar@1
  1342
     8 SI           NL     250.00000      50.00000     250.00000      235.32871       -.48520     289.09812 CU
alpar@1
  1343
                                            .48520     300.00000      255.06073          +Inf     298.67206 BIN3
alpar@1
  1344
\end{verbatim}
alpar@1
  1345
\end{scriptsize}
alpar@1
  1346
\end{landscape}
alpar@1
  1347
alpar@1
  1348
\begin{landscape}
alpar@1
  1349
\begin{scriptsize}
alpar@1
  1350
\begin{verbatim}
alpar@1
  1351
GLPK 4.42 - SENSITIVITY ANALYSIS REPORT                                                                         Page   2
alpar@1
  1352
alpar@1
  1353
Problem:    PLAN
alpar@1
  1354
Objective:  VALUE = 296.2166065 (MINimum)
alpar@1
  1355
alpar@1
  1356
   No. Column name  St      Activity      Obj coef   Lower bound       Activity      Obj coef  Obj value at Limiting
alpar@1
  1357
                                          Marginal   Upper bound          range         range   break point variable
alpar@1
  1358
------ ------------ -- ------------- ------------- -------------  ------------- ------------- ------------- ------------
alpar@1
  1359
     1 BIN1         NL        .             .03000        .           -28.82475       -.22362     288.90594 BIN4
alpar@1
  1360
                                            .25362     200.00000       33.88040          +Inf     304.80951 BIN4
alpar@1
  1361
alpar@1
  1362
     2 BIN2         BS     665.34296        .08000        .           802.22222        .01722     254.44822 BIN1
alpar@1
  1363
                                            .         2500.00000      313.43066        .08863     301.95652 MN
alpar@1
  1364
alpar@1
  1365
     3 BIN3         BS     490.25271        .17000     400.00000      788.61314        .15982     291.22807 MN
alpar@1
  1366
                                            .          800.00000     -347.42857        .17948     300.86548 BIN5
alpar@1
  1367
alpar@1
  1368
     4 BIN4         BS     424.18773        .12000     100.00000      710.52632        .10899     291.54745 MN
alpar@1
  1369
                                            .          700.00000     -256.15524        .14651     307.46010 BIN1
alpar@1
  1370
alpar@1
  1371
     5 BIN5         NL        .             .15000        .          -201.78739        .13544     293.27940 BIN3
alpar@1
  1372
                                            .01456    1500.00000       58.79586          +Inf     297.07244 BIN3
alpar@1
  1373
alpar@1
  1374
     6 ALUM         BS     299.63899        .21000        .           358.26772        .18885     289.87879 AL
alpar@1
  1375
                                            .               +Inf      112.40876        .22622     301.07527 MN
alpar@1
  1376
alpar@1
  1377
     7 SILICON      BS     120.57762        .38000        .           124.27093        .14828     268.27586 BIN5
alpar@1
  1378
                                            .               +Inf       85.54745        .46667     306.66667 MN
alpar@1
  1379
alpar@1
  1380
End of report
alpar@1
  1381
\end{verbatim}
alpar@1
  1382
\end{scriptsize}
alpar@1
  1383
\end{landscape}
alpar@1
  1384
alpar@1
  1385
\noindent
alpar@1
  1386
{\tt Marginal} is the reduced cost (dual activity) of the auxiliary
alpar@1
  1387
variable (row) or structural variable (column).
alpar@1
  1388
alpar@1
  1389
\medskip
alpar@1
  1390
alpar@1
  1391
\noindent
alpar@1
  1392
{\tt Lower bound} is the lower right-hand side (row) or lower bound
alpar@1
  1393
(column). If the row or column has no lower bound, this field contains
alpar@1
  1394
{\tt -Inf}.
alpar@1
  1395
alpar@1
  1396
\medskip
alpar@1
  1397
alpar@1
  1398
\noindent
alpar@1
  1399
{\tt Upper bound} is the upper right-hand side (row) or upper bound
alpar@1
  1400
(column). If the row or column has no upper bound, this field contains
alpar@1
  1401
{\tt +Inf}.
alpar@1
  1402
alpar@1
  1403
\bigskip
alpar@1
  1404
alpar@1
  1405
\noindent
alpar@1
  1406
{\it Sensitivity analysis of active bounds}
alpar@1
  1407
alpar@1
  1408
\medskip
alpar@1
  1409
alpar@1
  1410
\noindent
alpar@1
  1411
The sensitivity analysis of active bounds is performed only for rows,
alpar@1
  1412
which are active constraints, and only for non-basic columns, because
alpar@1
  1413
inactive constraints and basic columns have no active bounds.
alpar@1
  1414
alpar@1
  1415
For every auxiliary (row) or structural (column) non-basic variable the
alpar@1
  1416
routine starts changing its active bound in both direction. The first
alpar@1
  1417
of the two lines in the report corresponds to decreasing, and the
alpar@1
  1418
second line corresponds to increasing of the active bound. Since the
alpar@1
  1419
variable being analyzed is non-basic, its activity, which is equal to
alpar@1
  1420
its active bound, also starts changing. This changing leads to changing
alpar@1
  1421
of basic (auxiliary and structural) variables, which depend on the
alpar@1
  1422
non-basic variable. The current basis remains primal feasible and
alpar@1
  1423
therefore optimal while values of all basic variables are primal
alpar@1
  1424
feasible, i.e. are within their bounds. Therefore, if some basic
alpar@1
  1425
variable called the {\it limiting variable} reaches its (lower or
alpar@1
  1426
upper) bound first, before any other basic variables, it thereby limits
alpar@1
  1427
further changing of the non-basic variable, because otherwise the
alpar@1
  1428
current basis would become primal infeasible. The point, at which this
alpar@1
  1429
happens, is called the {\it break point}. Note that there are two break
alpar@1
  1430
points: the lower break point, which corresponds to decreasing of the
alpar@1
  1431
non-basic variable, and the upper break point, which corresponds to
alpar@1
  1432
increasing of the non-basic variable.
alpar@1
  1433
alpar@1
  1434
In the analysis report values of the non-basic variable (i.e. of its
alpar@1
  1435
active bound) being analyzed at both lower and upper break points are
alpar@1
  1436
printed in the field `{\tt Activity range}'. Corresponding values of
alpar@1
  1437
the objective function are printed in the field `{\tt Obj value at
alpar@1
  1438
break point}', and symbolic names of corresponding limiting basic
alpar@1
  1439
variables are printed in the field `{\tt Limiting variable}'.
alpar@1
  1440
If the active bound can decrease or/and increase unlimitedly, the field
alpar@1
  1441
`{\tt Activity range}' contains {\tt -Inf} or/and {\tt +Inf}, resp.
alpar@1
  1442
alpar@1
  1443
For example (see the example report above), row SI is a double-sided
alpar@1
  1444
constraint, which is active on its lower bound (right-hand side), and
alpar@1
  1445
its activity in the optimal solution being equal to the lower bound is
alpar@1
  1446
250. The activity range for this row is $[235.32871,255.06073]$. This
alpar@1
  1447
means that the basis remains optimal while the lower bound is
alpar@1
  1448
increasing up to 255.06073, and further increasing is limited by
alpar@1
  1449
(structural) variable BIN3. If the lower bound reaches this upper break
alpar@1
  1450
point, the objective value becomes equal to 298.67206.
alpar@1
  1451
alpar@1
  1452
Note that if the basis does not change, the objective function depends
alpar@1
  1453
on the non-basic variable linearly, and the per-unit change of the
alpar@1
  1454
objective function is the reduced cost (marginal value) of the
alpar@1
  1455
non-basic variable.
alpar@1
  1456
alpar@1
  1457
\bigskip
alpar@1
  1458
alpar@1
  1459
\noindent
alpar@1
  1460
{\it Sensitivity analysis of objective coefficients at non-basic
alpar@1
  1461
variables}
alpar@1
  1462
alpar@1
  1463
\medskip
alpar@1
  1464
alpar@1
  1465
\noindent
alpar@1
  1466
The sensitivity analysis of the objective coefficient at a non-basic
alpar@1
  1467
variable is quite simple, because in this case change in the objective
alpar@1
  1468
coefficient leads to equivalent change in the reduced cost (marginal
alpar@1
  1469
value).
alpar@1
  1470
alpar@1
  1471
For every auxiliary (row) or structural (column) non-basic variable the
alpar@1
  1472
routine starts changing its objective coefficient in both direction.
alpar@1
  1473
(Note that auxiliary variables are not included in the objective
alpar@1
  1474
function and therefore always have zero objective coefficients.) The
alpar@1
  1475
first of the two lines in the report corresponds to decreasing, and the
alpar@1
  1476
second line corresponds to increasing of the objective coefficient.
alpar@1
  1477
This changing leads to changing of the reduced cost of the non-basic
alpar@1
  1478
variable to be analyzed and does affect reduced costs of all other
alpar@1
  1479
non-basic variables. The current basis remains dual feasible and
alpar@1
  1480
therefore optimal while the reduced cost keeps its sign. Therefore, if
alpar@1
  1481
the reduced cost reaches zero, it limits further changing of the
alpar@1
  1482
objective coefficient (if only the non-basic variable is non-fixed).
alpar@1
  1483
alpar@1
  1484
In the analysis report minimal and maximal values of the objective
alpar@1
  1485
coefficient, on which the basis remains optimal, are printed in the
alpar@1
  1486
field `\verb|Obj coef range|'. If the objective coefficient can
alpar@1
  1487
decrease or/and increase unlimitedly, this field contains {\tt -Inf}
alpar@1
  1488
or/and {\tt +Inf}, resp.
alpar@1
  1489
alpar@1
  1490
For example (see the example report above), column BIN5 is non-basic
alpar@1
  1491
having its lower bound active. Its objective coefficient is 0.15, and
alpar@1
  1492
reduced cost in the optimal solution 0.01456. The column lower bound
alpar@1
  1493
remains active while the column reduced cost remains non-negative,
alpar@1
  1494
thus, minimal value of the objective coefficient, on which the current
alpar@1
  1495
basis still remains optimal, is $0.15-0.01456=0.13644$, that is
alpar@1
  1496
indicated in the field `\verb|Obj coef range|'.
alpar@1
  1497
alpar@1
  1498
\bigskip
alpar@1
  1499
alpar@1
  1500
\noindent
alpar@1
  1501
{\it Sensitivity analysis of objective coefficients at basic variables}
alpar@1
  1502
alpar@1
  1503
\medskip
alpar@1
  1504
alpar@1
  1505
\noindent
alpar@1
  1506
To perform sensitivity analysis for every auxiliary (row) or structural
alpar@1
  1507
(column) variable the routine starts changing its objective coefficient
alpar@1
  1508
in both direction. (Note that auxiliary variables are not included in
alpar@1
  1509
the objective function and therefore always have zero objective
alpar@1
  1510
coefficients.) The first of the two lines in the report corresponds to
alpar@1
  1511
decreasing, and the second line corresponds to increasing of the
alpar@1
  1512
objective coefficient. This changing leads to changing of reduced costs
alpar@1
  1513
of non-basic variables. The current basis remains dual feasible and
alpar@1
  1514
therefore optimal while reduced costs of all non-basic variables
alpar@1
  1515
(except fixed variables) keep their signs. Therefore, if the reduced
alpar@1
  1516
cost of some non-basic non-fixed variable called the {\it limiting
alpar@1
  1517
variable} reaches zero first, before reduced cost of any other
alpar@1
  1518
non-basic non-fixed variable, it thereby limits further changing of the
alpar@1
  1519
objective coefficient, because otherwise the current basis would become
alpar@1
  1520
dual infeasible (non-optimal). The point, at which this happens, is
alpar@1
  1521
called the {\it break point}. Note that there are two break points: the
alpar@1
  1522
lower break point, which corresponds to decreasing of the objective
alpar@1
  1523
coefficient, and the upper break point, which corresponds to increasing
alpar@1
  1524
of the objective coefficient. Let the objective coefficient reach its
alpar@1
  1525
limit value and continue changing a bit further in the same direction
alpar@1
  1526
that makes the current basis dual infeasible (non-optimal). Then the
alpar@1
  1527
reduced cost of the non-basic limiting variable becomes ``a bit'' dual
alpar@1
  1528
infeasible that forces the limiting variable to enter the basis
alpar@1
  1529
replacing there some basic variable, which leaves the basis to keep its
alpar@1
  1530
primal feasibility. It should be understood that if we change the
alpar@1
  1531
current basis in this way exactly at the break point, both the current
alpar@1
  1532
and adjacent bases will be optimal with the same objective value,
alpar@1
  1533
because at the break point the limiting variable has zero reduced cost.
alpar@1
  1534
On the other hand, in the adjacent basis the value of the limiting
alpar@1
  1535
variable changes, because there it becomes basic, that leads to
alpar@1
  1536
changing of the value of the basic variable being analyzed. Note that
alpar@1
  1537
on determining the adjacent basis the bounds of the analyzed basic
alpar@1
  1538
variable are ignored as if it were a free (unbounded) variable, so it
alpar@1
  1539
cannot leave the current basis.
alpar@1
  1540
alpar@1
  1541
In the analysis report lower and upper limits of the objective
alpar@1
  1542
coefficient at the basic variable being analyzed, when the basis
alpar@1
  1543
remains optimal, are printed in the field `{\tt Obj coef range}'.
alpar@1
  1544
Corresponding values of the objective function at both lower and upper
alpar@1
  1545
break points are printed in the field `{\tt Obj value at break point}',
alpar@1
  1546
symbolic names of corresponding non-basic limiting variables are
alpar@1
  1547
printed in the field `{\tt Limiting variable}', and values of the basic
alpar@1
  1548
variable, which it would take on in the adjacent bases (as was
alpar@1
  1549
explained above) are printed in the field `{\tt Activity range}'.
alpar@1
  1550
If the objective coefficient can increase or/and decrease unlimitedly,
alpar@1
  1551
the field `{\tt Obj coef range}' contains {\tt -Inf} and/or {\tt +Inf},
alpar@1
  1552
resp. It also may happen that no dual feasible adjacent basis exists
alpar@1
  1553
(i.e. on entering the basis the limiting variable can increase or
alpar@1
  1554
decrease unlimitedly), in which case the field `{\tt Activity range}'
alpar@1
  1555
contains {\tt -Inf} and/or {\tt +Inf}.
alpar@1
  1556
alpar@1
  1557
\newpage
alpar@1
  1558
alpar@1
  1559
For example (see the example report above), structural variable
alpar@1
  1560
(column) BIN3 is basic, its optimal value is 490.25271, and its
alpar@1
  1561
objective coefficient is 0.17. The objective coefficient range for this
alpar@1
  1562
column is $[0.15982,0.17948]$. This means that the basis remains
alpar@1
  1563
optimal while the objective coefficient is decreasing down to 0.15982,
alpar@1
  1564
and further decreasing is limited by (auxiliary) variable MN. If we
alpar@1
  1565
make the objective coefficient a bit less than 0.15982, the limiting
alpar@1
  1566
variable MN will enter the basis, and in that adjacent basis the
alpar@1
  1567
structural variable BIN3 will take on new optimal value 788.61314. At
alpar@1
  1568
the lower break point, where the objective coefficient is exactly
alpar@1
  1569
0.15982, the objective function takes on the value 291.22807 in both
alpar@1
  1570
the current and adjacent bases.
alpar@1
  1571
alpar@1
  1572
Note that if the basis does not change, the objective function depends
alpar@1
  1573
on the objective coefficient at the basic variable linearly, and the
alpar@1
  1574
per-unit change of the objective function is the value of the basic
alpar@1
  1575
variable.
alpar@1
  1576
alpar@1
  1577
%* eof *%