数学建模社区-数学中国

标题: 讲讲官方的例子，MATLAB调用Lingo [打印本页]

作者: wujianjack2 时间: 2015-4-5 19:22
标题: 讲讲官方的例子，MATLAB调用Lingo
以我的习惯，在MATLAB安装目录下新建一个名为works的文件夹，将Staff.m和STAFFPTR.lng文件放在这个文件目录下，然后把Lingd15.h和Lingd15.lib放在LINGO安装目录下，好，开始吧。
1. STAFFPTR.lng
MODEL:
SETS:
DAYS / MON TUE WED THU FRI SAT SUN/:
   NEEDS, START, ONDUTY;
ENDSETS

[OBJECTIVE] MIN = @SUM( DAYS( I): START( I));

@FOR( DAYS( TODAY):

! Calculate number on duty;
ONDUTY( TODAY) =
   @SUM( DAYS( D)| D #LE# 5:
   START( @WRAP( TODAY - D + 1, @SIZE( DAYS))));

! Enforce staffing requirement;
ONDUTY( TODAY) >= NEEDS( TODAY);

@GIN( START);
);

DATA:
NEEDS =  @POINTER( 1);
@POINTER( 2) = START;
@POINTER( 3) = ONDUTY;
@POINTER( 4) = OBJECTIVE;
@POINTER( 5) = @STATUS();
ENDDATA
END

2. Staff.m
clear; clc;

% By default, Lingo API is in the d:\LINGO15 folder
addpath('D:\\LINGO15');

% Load the Lingo API
loadlibrary('Lingd15', 'Lingd15.h');
if ~libisloaded('Lingd15')
error('Cannot load LINGO library');
end

% Print out LINGO library functions
libfunctions('Lingd15', '-full');

% Create the Lingo environment
pLingo = calllib('Lingd15', 'LScreateEnvLng');
if ~( exist('pLingo', 'var') && isa(pLingo, 'lib.pointer') )
error('Cannot create LINGO environment object');
end

% Open a Lingo log file
nError = calllib('Lingd15', 'LSopenLogFileLng', pLingo, 'D:\\LINGO15\\testLingoLib15.log');
if ~strcmp( nError, 'LSERR_NO_ERROR_LNG')
error('Cannot open LINGO log file');
end

% Staffing needs
dNeeds = [1 4 3 7 2 5 3];

% Pointers to memory transfer areas
dNeedsPtr = libpointer('doublePtr', dNeeds);
dStartPtr = libpointer('doublePtr', zeros(1, 7));
dOnDutyPtr = libpointer('doublePtr', zeros(1, 7));
dStatusPtr = libpointer('doublePtr', -1);
dTotalPtr = libpointer('doublePtr', 0);
nPointerNowPtr = libpointer('int32Ptr', 0);

% Pass memory transfer pointers to LINGO
nError = calllib('Lingd15', 'LSsetPointerLng', pLingo, dNeedsPtr, nPointerNowPtr);
if ~strcmp( nError, 'LSERR_NO_ERROR_LNG')
error('Cannot assign pointer');
end
nError = calllib('Lingd15', 'LSsetPointerLng', pLingo, dStartPtr, nPointerNowPtr);
if ~strcmp( nError, 'LSERR_NO_ERROR_LNG')
error('Cannot assign pointer');
end
nError = calllib('Lingd15', 'LSsetPointerLng', pLingo, dOnDutyPtr, nPointerNowPtr);
if ~strcmp( nError, 'LSERR_NO_ERROR_LNG')
error('Cannot assign pointer');
end
nError = calllib('Lingd15', 'LSsetPointerLng', pLingo, dTotalPtr, nPointerNowPtr);
if ~strcmp( nError, 'LSERR_NO_ERROR_LNG')
error('Cannot assign pointer');
end
nError = calllib('Lingd15', 'LSsetPointerLng', pLingo, dStatusPtr, nPointerNowPtr);
if ~strcmp( nError, 'LSERR_NO_ERROR_LNG')
error('Cannot assign pointer');
end

% Set up the command script
csScript = 'SET ECHOIN 1 \n';
csScript = [csScript 'TAKE STAFFPTR.lng \n'];
csScript = [csScript 'GO \n'];
csScript = [csScript 'QUIT \n'];

% Run the script
nError = calllib('Lingd15', 'LSexecuteScriptLng', pLingo, sprintf(csScript));
if ~strcmp( nError, 'LSERR_NO_ERROR_LNG')
error('Cannot execute script');
end

% Close log file
calllib('Lingd15', 'LScloseLogFileLng', pLingo);

% Optimal?
if get(dStatusPtr, 'value')
error('Unable to solve!');
else
% Display solution
disp(['Start  : ' num2str(dStartPtr.value)]);
disp(['Needs  : ' num2str(dNeedsPtr.value)]);
disp(['On Duty: ' num2str(dOnDutyPtr.value)]);
end

% Free Lingo environment
calllib('Lingd15', 'LSdeleteEnvLng', pLingo);

3. 注意，MATLAB当前目录为先前创建的works文件夹，运行看结果：
MATLAB显示的：
Warning: The data type 'FcnPtr' used by function LSsetCallbackErrorLng does not exist.
> In loadlibrary (line 431)
  In Staff (line 7)
Warning: The data type 'FcnPtr' used by function LSsetCallbackSolverLng does not exist.
> In loadlibrary (line 431)
  In Staff (line 7)
Start  : 0  4  1  0  0  0  2
Needs  : 1  4  3  7  2  5  3
On Duty: 2  6  7  7  5  5  3

testLingoLib15.log文件中的：
  Global optimal solution found.
  Objective value:                      7.000000000000
  Objective bound:                      7.000000000000
  Infeasibilities:                      0.000000000000
  Extended solver steps:                            0
  Total solver iterations:                            4
  Elapsed runtime seconds:                         0.00

  Running output operations ...

  Model Class:                                     MILP

  Total variables:                   14
  Nonlinear variables:                0
  Integer variables:                   7

  Total constraints:                15
  Nonlinear constraints:             0

  Total nonzeros:                   56
  Nonlinear nonzeros:                0

                  Variable                Value             Reduced Cost
               NEEDS( MON)       1.000000000000          0.000000000000
               NEEDS( TUE)       4.000000000000          0.000000000000
               NEEDS( WED)       3.000000000000          0.000000000000
               NEEDS( THU)       7.000000000000          0.000000000000
               NEEDS( FRI)       2.000000000000          0.000000000000
               NEEDS( SAT)       5.000000000000          0.000000000000
               NEEDS( SUN)       3.000000000000          0.000000000000
               START( MON)       0.000000000000          1.000000000000
               START( TUE)       4.000000000000          1.000000000000
               START( WED)       1.000000000000          1.000000000000
               START( THU)       0.000000000000          1.000000000000
               START( FRI)       0.000000000000          1.000000000000
               START( SAT)       0.000000000000          1.000000000000
               START( SUN)       2.000000000000          1.000000000000
            ONDUTY( MON)       2.000000000000          0.000000000000
            ONDUTY( TUE)       6.000000000000          0.000000000000
            ONDUTY( WED)       7.000000000000          0.000000000000
            ONDUTY( THU)       7.000000000000          0.000000000000
            ONDUTY( FRI)       5.000000000000          0.000000000000
            ONDUTY( SAT)       5.000000000000          0.000000000000
            ONDUTY( SUN)       3.000000000000          0.000000000000

                     Row       Slack or Surplus          Dual Price
               OBJECTIVE       7.000000000000          -1.000000000000
                        2       0.000000000000          0.000000000000
                        3       1.000000000000          0.000000000000
                        4       0.000000000000          0.000000000000
                        5       2.000000000000          0.000000000000
                        6       0.000000000000          0.000000000000
                        7       4.000000000000          0.000000000000
                        8       0.000000000000          0.000000000000
                        9       0.000000000000          0.000000000000
                        10       0.000000000000          0.000000000000
                        11       3.000000000000          0.000000000000
                        12       0.000000000000          0.000000000000
                        13       0.000000000000          0.000000000000
                        14       0.000000000000          0.000000000000
                        15       0.000000000000          0.000000000000

: QUIT

OK, done!

作者: beifang 时间: 2015-4-8 08:33
请问，你装的matlab和lingo15.0是多少位的了？我运行怎么出错，说找不到文件Lingd15.h呢？

作者: beifang 时间: 2015-4-8 08:43
运行出错提示如下：

索引超出矩阵维度。

出错
loadlibrary>getLoadlibraryCompilerConfiguration
(line 497)
      compilerConfiguration=compilerConfiguration(1);
      %unix machines return c and cpp
      compilers here

出错 loadlibrary (line 253)
[thunk_build_fn,preprocess_command]=getLoadlibraryCompilerConfiguration(ccinclude,header,headername,compilerConfiguration);

出错 Staff (line 7)
loadlibrary('Lingd15', 'Lingd15.h');

欢迎光临数学建模社区-数学中国 (http://www.madio.net/)