path: root/build/Scilab/intfmincon.sci

// Copyright (C) 2015 - IIT Bombay - FOSSEE
//
// This file must be used under the terms of the CeCILL.
// This source file is licensed as described in the file COPYING, which
// you should have received as part of this distribution.  The terms
// are also available at
// http://www.cecill.info/licences/Licence_CeCILL_V2-en.txt
// Author: Harpreet Singh, Pranav Deshpande and Akshay Miterani
// Organization: FOSSEE, IIT Bombay
// Email: toolbox@scilab.in

function [xopt,fopt,exitflag,gradient,hessian] = intfmincon (varargin)
  // Solves a constrainted multi-variable mixed integer non linear programming problem
  //
  //   Calling Sequence
  //   xopt = intfmincon(f,x0,intcon,A,b)
  //   xopt = intfmincon(f,x0,intcon,A,b,Aeq,beq)
  //   xopt = intfmincon(f,x0,intcon,A,b,Aeq,beq,lb,ub)
  //   xopt = intfmincon(f,x0,intcon,A,b,Aeq,beq,lb,ub,nlc)
  //   xopt = intfmincon(f,x0,intcon,A,b,Aeq,beq,lb,ub,nlc,options)
  //   [xopt,fopt] = intfmincon(.....)
  //   [xopt,fopt,exitflag]= intfmincon(.....)
  //   [xopt,fopt,exitflag,gradient]=intfmincon(.....)
  //   [xopt,fopt,exitflag,gradient,hessian]=intfmincon(.....)
	//
	//   Parameters
	//   f : a function, representing the objective function of the problem 
  //   x0 : a vector of doubles, containing the starting values of variables. 
  //   intcon : a vector of integers, represents which variables are constrained to be integers
  //   A : a matrix of double, represents the linear coefficients in the inequality constraints A⋅x ≤ b. 
  //   b : a vector of double, represents the linear coefficients in the inequality constraints A⋅x ≤ b.
  //   Aeq : a matrix of double, represents the linear coefficients in the equality constraints Aeq⋅x = beq.
  //   beq : a vector of double, represents the linear coefficients in the equality constraints Aeq⋅x = beq.
  //   lb : Lower bounds, specified as a vector or array of double. lb represents the lower bounds elementwise in lb ≤ x ≤ ub.
  //   ub : Upper bounds, specified as a vector or array of double. ub represents the upper bounds elementwise in lb ≤ x ≤ ub.
  //   nlc : a function, representing the Non-linear Constraints functions(both Equality and Inequality) of the problem. It is declared in such a way that non-linear inequality constraints are defined first as a single row vector (c), followed by non-linear equality constraints as another single row vector (ceq). Refer Example for definition of Constraint function.
  //   options : a list, containing the option for user to specify. See below for details. 
  //   xopt : a vector of doubles, containing the the computed solution of the optimization problem.
  //   fopt : a scalar of double, containing the the function value at x.
  //   exitflag : a scalar of integer, containing the flag which denotes the reason for termination of algorithm. See below for details.
  //   gradient : a vector of doubles, containing the Objective's gradient of the solution.
  //   hessian  : a matrix of doubles, containing the Objective's hessian of the solution.
  //
	//   Description
	//   Search the minimum of a multi-variable function on bounded interval specified by :
	//   Find the minimum of f(x) such that 
	//
	//   <latex>
	//    \begin{eqnarray}
	//    &\mbox{min}_{x}
	//    & f(x)\\
	//    & \text{subject to} & x1 \ < x \ < x2 \\
	//    \end{eqnarray}
	//   </latex>
	//
	//   The routine calls Bonmin for solving the Bounded Optimization problem, Bonmin is a library written in C++.
	//
	// The options allows the user to set various parameters of the Optimization problem. 
	// It should be defined as type "list" and contains the following fields.
	// <itemizedlist>
	//   <listitem>Syntax : options= list("IntegerTolerance", [---], "MaxNodes",[---], "MaxIter", [---], "AllowableGap",[---] "CpuTime", [---],"gradobj", "off", "hessian", "off" );</listitem>
  //   <listitem>IntegerTolerance : a Scalar, a number with that value of an integer is considered integer..</listitem>
  //   <listitem>MaxNodes : a Scalar, containing the Maximum Number of Nodes that the solver should search.</listitem>
	//   <listitem>CpuTime : a Scalar, containing the Maximum amount of CPU Time that the solver should take.</listitem>
	//   <listitem>AllowableGap : a Scalar, to stop the tree search when the gap between the objective value of the best known solution is reached.</listitem>
  //   <listitem>MaxIter : a Scalar, containing the Maximum Number of Iteration that the solver should take.</listitem>
  //   <listitem>gradobj : a string, to turn on or off the user supplied objective gradient.</listitem>
  //   <listitem>hessian : a Scalar, to turn on or off the user supplied objective hessian.</listitem>
	//   <listitem>Default Values : options = list('integertolerance',1d-06,'maxnodes',2147483647,'cputime',1d10,'allowablegap',0,'maxiter',2147483647,'gradobj',"off",'hessian',"off")</listitem>
	// </itemizedlist>
	//
	// The exitflag allows to know the status of the optimization which is given back by Ipopt.
	// <itemizedlist>
	//   <listitem>exitflag=0 : Optimal Solution Found </listitem>
	//   <listitem>exitflag=1 : Maximum Number of Iterations Exceeded. Output may not be optimal.</listitem>
	//   <listitem>exitflag=2 : Maximum CPU Time exceeded. Output may not be optimal.</listitem>
	//   <listitem>exitflag=3 : Stop at Tiny Step.</listitem>
	//   <listitem>exitflag=4 : Solved To Acceptable Level.</listitem>
	//   <listitem>exitflag=5 : Converged to a point of local infeasibility.</listitem>
	// </itemizedlist>
	//
	// For more details on exitflag see the Bonmin documentation, go to http://www.coin-or.org/Bonmin
  //
	// Examples
	//	//Find x in R^6 such that it minimizes:
	//    //f(x)= sin(x1) + sin(x2) + sin(x3) + sin(x4) + sin(x5) + sin(x6)
	//	//-2 <= x1,x2,x3,x4,x5,x6 <= 2
	//    //Objective function to be minimised
	//    function y=f(x)
	//		y=0
	//		for i =1:6
	//			y=y+sin(x(i));
	//		end	
	//	endfunction
	//	//Variable bounds  
	//	x1 = [-2, -2, -2, -2, -2, -2];
	//  x2 = [2, 2, 2, 2, 2, 2];
  //  intcon = [2 3 4]
	//	//Options
	//	options=list("MaxIter",[1500],"CpuTime", [100])
	//	[x,fval] =intfmincon(f ,intcon, x1, x2, options)
  // // Press ENTER to continue
	//
	// Examples
	//	//Find x in R such that it minimizes:
	//    //f(x)= 1/x^2
	//	//0 <= x <= 1000
	//    //Objective function to be minimised
	//    function y=f(x)
	//		   y=1/x^2;
	//	  endfunction
	//	//Variable bounds  
	//	x1 = [0];
	//  x2 = [1000];
  //  intcon = [1];
	//	[x,fval,exitflag,output,lambda] =intfmincon(f,intcon , x1, x2)
  // // Press ENTER to continue
	//
	// Examples
	//    //The below problem is an unbounded problem:
	//	//Find x in R^2 such that it minimizes:
	//    //f(x)= -[(x1-1)^2 + (x2-1)^2]
	//	//-inf <= x1,x2 <= inf
	//    //Objective function to be minimised
	//    function y=f(x)
	// 		y=-((x(1)-1)^2+(x(2)-1)^2);
	//	endfunction
	//	//Variable bounds  
	//	x1 = [-%inf , -%inf];
	//  x2 = [ %inf , %inf];
	//	//Options
	//	options=list("MaxIter",[1500],"CpuTime", [100])
	//	[x,fval,exitflag,output,lambda] =intfmincon(f,intcon, x1, x2, options)  
	// Authors
	// Harpreet Singh

	//To check the number of input and output arguments
 	[lhs , rhs] = argn();
	
	//To check the number of arguments given by the user
 	if ( rhs<4 | rhs>11 ) then
  		errmsg = msprintf(gettext("%s: Unexpected number of input arguments : %d provided while should be int [4 5] "), "intfmincon", rhs);
  		error(errmsg);
 	end
 
	//Storing the Input Parameters  
 	fun = varargin(1);
 	x0     = varargin(2);
  intcon = varargin(3); 
  A      = varargin(4);
  b      = varargin(5);
  Aeq    = [];
  beq    = [];
  lb       = [];
  ub       = [];
  nlc      = [];
    
  if (rhs>5) then
    Aeq    = varargin(6);
    beq    = varargin(7);
  end

  if (rhs>7) then
    lb       = varargin(8);
    ub       = varargin(9);
  end

  if (rhs>9) then
    nlc      = varargin(10);
  end

  param = list();
  //To check whether options has been entered by user   
  if ( rhs>  10) then
      param =varargin(11);
  end

  //To check whether the Input arguments
  Checktype("intfmincon", fun, "fun", 1, "function");
  Checktype("intfmincon", x0, "x0", 2, "constant");
  Checktype("intfmincon", intcon, "intcon", 3, "constant");
  Checktype("intfmincon", A, "A", 4, "constant");
  Checktype("intfmincon", b, "b", 5, "constant");
  Checktype("intfmincon", Aeq, "Aeq", 6, "constant");
  Checktype("intfmincon", beq, "beq", 7, "constant");
  Checktype("intfmincon", lb, "lb", 8, "constant");
  Checktype("intfmincon", ub, "ub", 9, "constant");
  Checktype("intfmincon", nlc, "nlc", 10, ["constant","function"]);
  Checktype("intfmincon", param, "options", 11, "list");  


  nbVar = size(x0,"*");
  if(nbVar==0) then
    errmsg = msprintf(gettext("%s: x0 cannot be an empty"), "intfmincon");
    error(errmsg);    
  end

  if(size(lb,"*")==0) then
    lb = repmat(-%inf,nbVar,1);
  end

  if(size(ub,"*")==0) then
    ub = repmat(%inf,nbVar,1);
  end

  //////////////// To Check linear constraints /////////

  //To check for correct size of A(3rd paramter)
  if(size(A,2)~=nbVar & size(A,2)~=0) then
    errmsg = msprintf(gettext("%s: Expected Matrix of size (No of linear inequality constraints X No of Variables) or an Empty Matrix for Linear Inequality Constraint coefficient Matrix A"), solver_name);
    error(errmsg);
  end
  nbConInEq=size(A,"r");

  //To check for the correct size of Aeq (5th paramter)
  if(size(Aeq,2)~=nbVar & size(Aeq,2)~=0) then
    errmsg = msprintf(gettext("%s: Expected Matrix of size (No of linear equality constraints X No of Variables) or an Empty Matrix for Linear Equality Constraint coefficient Matrix Aeq"), solver_name);
    error(errmsg);
  end
  nbConEq=size(Aeq,"r");

  ///////////////// To check vectors ///////////////// 

  Checkvector("intfmincon", x0, "x0", 2, nbVar);
  x0 = x0(:);
  if(size(intcon,"*")) then
    Checkvector("intfmincon", intcon, "intcon", 3, size(intcon,"*"))
    intcon = intcon(:);
  end
  if(nbConInEq) then
    Checkvector("intfmincon", b, "b", 5, nbConInEq);
    b = b(:);
  end
  if(nbConEq) then
    Checkvector("intfmincon", beq, "beq", 7, nbConEq);
    beq = beq(:);
  end
  Checkvector("intfmincon", lb, "lb", 8, nbVar);
  lb = lb(:);
  
  Checkvector("intfmincon", ub, "ub", 9, nbVar);
  ub = ub(:);

  /////////////// To check integer //////////////////////
  for i=1:size(intcon,1)
      if(intcon(i)>nbVar) then
        errmsg = msprintf(gettext("%s: The values inside intcon should be less than the number of variables"), "intfmincon");
        error(errmsg);
      end

      if (intcon(i)<0) then
        errmsg = msprintf(gettext("%s: The values inside intcon should be greater than 0 "), "intfmincon");
        error(errmsg);
      end

      if(modulo(intcon(i),1)) then
        errmsg = msprintf(gettext("%s: The values inside intcon should be an integer "), "intfmincon");
        error(errmsg);
      end
  end
  
options = list('integertolerance',1d-06,'maxnodes',2147483647,'cputime',1d10,'allowablegap',0,'maxiter',2147483647,'gradobj',"off",'hessian',"off",'gradcon',"off")

  //Pushing param into default value
  
  for i = 1:(size(param))/2
    select convstr(param(2*i-1),'l')
      case 'integertolerance' then
        Checktype("intfmincon_options", param(2*i), param(2*i-1), 2*i, "constant");
        options(2) = param(2*i);
      case 'maxnodes' then
        Checktype("intfmincon_options", param(2*i), param(2*i-1), 2*i, "constant");
        options(4) = param(2*i);
      case 'cputime' then 
        Checktype("intfmincon_options", param(2*i), param(2*i-1), 2*i, "constant");
        options(6) = param(2*i);
      case 'allowablegap' then
        Checktype("intfmincon_options", param(2*i), param(2*i-1), 2*i, "constant");
        options(8) = param(2*i);
      case 'maxiter' then
        Checktype("intfmincon_options", param(2*i), param(2*i-1), 2*i, "constant");
        options(10) = param(2*i);
      case 'gradobj' then
        Checktype("intfmincon_options", param(2*i), param(2*i-1), 2*i, "string");
        if(convstr(options(2*i),'l') == "on") then
          options(12) = "on"
        elseif(convstr(options(2*i),'l') == "off") then
          options(12) = "off"
        else
          error(999, 'Unknown string passed in gradobj.');
        end
      case 'hessian' then
        Checktype("intfmincon_options", param(2*i), param(2*i-1), 2*i, "function");
        options(14) = param(2*i);
      case 'GradCon' then
        Checktype("intfmincon_options", param(2*i), param(2*i-1), 2*i, "string");
        if(convstr(options(2*i),'l') == "on") then
          options(16) = "on"
        elseif(convstr(options(2*i),'l') == "off") then
          options(16) = "off"
        else
          error(999, 'Unknown string passed in gradcon.');
        end
      else    
        error(999, 'Unknown string argument passed.');
    end
  end 

  ///////////////// Functions Check /////////////////

  //To check the match between f (1st Parameter) and x0 (2nd Parameter)
  if(execstr('init=fun(x0)','errcatch')==21) then
    errmsg = msprintf(gettext("%s: Objective function and x0 did not match"), "intfmincon");
    error(errmsg);
  end

  if(options(12) == "on") then
       if(execstr('[grad_y,grad_dy]=fun(x1)','errcatch')==59) then
        errmsg = msprintf(gettext("%s: Gradient of objective function is not provided"), "intfmincon");
        error(errmsg);
      end
      Checkvector("intfmincon_options", grad_dy, "dy", 12, nbVar);
  end

  if(options(14) == "on") then
      if(execstr('[hessian_y,hessian_dy,hessian]=fun(x1)','errcatch')==59) then
        errmsg = msprintf(gettext("%s: Gradient of objective function is not provided"), "intfmincon");
        error(errmsg);
      end

      if ( ~isequal(size(hessian) == [nbVar nbVar]) ) then
        errmsg = msprintf(gettext("%s: Size of hessian should be nbVar X nbVar"), "intfmincon");
        error(errmsg);
      end
  end

  numNlic = 0;
  numNlec = 0;
  numNlc = 0;

  if (type(nlc) == 13 | type(nlc) == 11) then    
      if(execstr('[sample_c,sample_ceq] = nlc(x0)','errcatch')==21) then
        errmsg = msprintf(gettext("%s: Non-Linear Constraint function and x0 did not match"), solver_name);
        error(errmsg);
      end
      numNlic = size(sample_c,"*");
      numNlec = size(sample_ceq,"*");
      numNlc = no_nlic + no_nlec;
  end

  /////////////// Creating conLb and conUb ////////////////////////

  conLb = [repmat(-%inf,nbConInEq,1);beq;repmat(-%inf,numNlic,1);repmat(0,numNlic,1);]
  conUb = [b;beq;repmat(0,numNlic,1);repmat(0,numNlic,1);]

  //Converting the User defined Objective function into Required form (Error Detectable)
  function [y,check] = _f(x)
    try
      y=fun(x)
      [y,check] =  checkIsreal(y)
    catch
      y=0;
      check=1;
    end
   endfunction
    
   //Defining an inbuilt Objective gradient function 
    function [dy,check] = _gradf(x) 
      if (options(12) =="on") then
        try
          [y,dy]=fun(x)
          [dy,check] =  checkIsreal(dy)
        catch
          dy = 0;
          check=1;
        end
      else
          try 
            dy=numderivative(fun,x)
            [dy,check] =  checkIsreal(dy)
          catch
            dy=0;
          check=1;
          end
      end
    endfunction

    function [y,check] = _addnlc(x)
    x= x(:);
    c=[]
    ceq = [];
      try
        if(type(nlc) == 13 & numNlc~=0) then
          {
            [c,ceq] = nlc(x);
          }
        end
        ylin = [A*x;Aeq*x];
        y = [ylin;c(:);ceq(:)];
        [y,check] = checkIsreal(y)
      catch
        y=0;
        check=1;
      end
    endfunction

    //Defining an inbuilt Objective gradient function 
    function [dy,check] = _gradnlc(x) 
      if (options(16) =="on") then
        try
          [y,dy]=_addnlc(x)
          [dy,check] = checkIsreal(dy)
        catch
          dy = 0;
          check=1;
        end
      else
          try
            dy=numderivative(_addnlc,x)
            [dy,check] = checkIsreal(dy)
          catch
            dy=0;
            check=1;
          end
      end
    endfunction

  //Defining a function to calculate Hessian if the respective user entry is OFF 
  function [hessy,check]=_gradhess(x,obj_factor,lambda)
    if (type(options(14)) == "function") then
      try
        [obj,dy,hessy] = fun(x,obj_factor,lambda)
        [hessy,check] =  checkIsreal(hessy)
      catch
        hessy = 0;
        check=1;
      end
    else
      try
        [dy,hessfy]=numderivative(_f,x,%eps^(1/3),1,"blockmat");
        hessny = []
        if(type(nlc) == 13 & numNlc~=0) then
        {
          [dy,hessny] = numderivative(_addnlc,x,%eps^(1/3),1,"blockmat");
        }
        end
        hessianc = []
          for i = 1:numNlc
                    hessianc = hessianc + lambda(i)*hessny((i-1)*nbVar+1:nbVar*i,:)
          end
        hessy = obj_factor*hessfy + hessianc;
        [hessy,check] =  checkIsreal(hessy)
      catch
        hessy=0;
        check=1;
      end
    end
  endfunction

    intconsize = size(intcon,"*");

	[xopt,fopt,exitflag] = inter_fmincon(_f,_gradf,_addnlc,_gradnlc,_gradhess,x0,lb,ub,conLb,conUb,intcon,options,nbConInEq+nbConEq);

    //In the cases of the problem not being solved, return NULL to the output matrices
    if( exitflag~=0 & exitflag~=3 ) then
      gradient = [];
      hessian = [];
    else
      [ gradient, hessian] = numderivative(_f, xopt, [], [], "blockmat");
    end

    //To print output message
    select exitflag
    
    case 0 then
        printf("\nOptimal Solution Found.\n");
    case 1 then
        printf("\nInFeasible Solution.\n");
    case 2 then
        printf("\nObjective Function is Continuous Unbounded.\n");
    case 3 then
        printf("\Limit Exceeded.\n");
    case 4 then
        printf("\nUser Interrupt.\n");
    case 5 then
        printf("\nMINLP Error.\n");
    else
        printf("\nInvalid status returned. Notify the Toolbox authors\n");
        break;
    end
endfunction

function [y, check] = checkIsreal(x)
  if ((~isreal(x))) then
    y = 0
    check=1;
  else
        y = x;
    check=0;
  end 
endfunction