/* nag_opt_estimate_deriv (e04xac) Example Program.
 *
 * Copyright 1998 Numerical Algorithms Group.
 *
 * Mark 5, 1998.
 * Mark 7 revised, 2001.
 * Mark 8 revised, 2004.
 *
 */

#include <nag.h>
#include <nag_stdlib.h>
#include <stdio.h>
#include <nage04.h>

static int ex1(void);
static int ex2(void);
#ifdef __cplusplus
extern "C" {
#endif
  static void objfun(Integer n, double x[], double *objf,
                     double g[], Nag_Comm *comm);
#ifdef __cplusplus
}
#endif
static void objfun(Integer n, double x[], double *objf,
                   double g[], Nag_Comm *comm)
{
  double a, asq, b, bsq, c, csq, d, dsq;

  a = x[0] + 10.0*x[1];
  b = x[2] - x[3];
  c = x[1] - 2.0*x[2];
  d = x[0] - x[3];
  asq = a*a;
  bsq = b*b;
  csq = c*c;
  dsq = d*d;
  *objf = asq + 5.0*bsq + csq*csq + 10.0*dsq*dsq;
  if (comm->flag == 2)
    {
      g[0] =   2.0*a + 40.0*d*dsq;
      g[1] =  20.0*a + 4.0*c*csq;
      g[2] =  10.0*b -  8.0*c*csq;
      g[3] = -10.0*b - 40.0*d*dsq;

    }
}
/* objfun */

int main(void)
{

  Integer exit_status_ex1=0;
  Integer exit_status_ex2=0;

  Vprintf("nag_opt_estimate_deriv (e04xac) Example Program Results\n");

  exit_status_ex1 = ex1();
  exit_status_ex2 = ex2();

  return exit_status_ex1 == 0 && exit_status_ex2 == 0 ? 0 : 1;
}
#define H(I,J) h[(I)*tdh + J]

static int ex1(void)
{
#define MAXN 4

  /* Local variables */
  Integer exit_status=0, n, tdh;
  NagError fail;
  Nag_DerivInfo *deriv_info=0;
  double *g=0, *h=0, objf, *x=0;

  INIT_FAIL(fail);


  n = MAXN;
  if (n>=1)
    {
      if ( !( x = NAG_ALLOC(n, double)) ||
           !( g = NAG_ALLOC(n, double)) ||
           !( h = NAG_ALLOC(n*n, double)) ||
           !( deriv_info = NAG_ALLOC(n, Nag_DerivInfo))
           )
        {
          Vprintf("Allocation failure\n");
          exit_status = -1;
          goto END;
        }
      tdh = n;
    }
  else
    {
      Vprintf("Invalid n.\n");
      exit_status = 1;
      return exit_status;
    }
  Vprintf("\nExample 1: default options\n");

  x[0] = 3.0;
  x[1] = -1.0;
  x[2] = 0.0;
  x[3] = 1.0;

  /* Pass null pointers for the h_central and h_forward parameters
   * as we do not need these values.
   */
  /* nag_opt_estimate_deriv (e04xac).
   * Computes an approximation to the gradient vector and/or
   * the Hessian matrix for use with nag_opt_nlp (e04ucc) and
   * other nonlinear optimization functions
   */
  nag_opt_estimate_deriv(n, x, objfun, &objf, g, (double*)0, (double*)0,
                         h, tdh, deriv_info, E04_DEFAULT, NAGCOMM_NULL, &fail);
  if (fail.code != NE_NOERROR)
    {
      Vprintf("Error from nag_opt_estimate_deriv (e04xac).\n%s\n",
              fail.message);
      exit_status = 1;
      goto END;
    }
 END:
  if (x) NAG_FREE(x);
  if (g) NAG_FREE(g);
  if (h) NAG_FREE(h);
  if (deriv_info) NAG_FREE(deriv_info);
  return exit_status;
} /* ex1 */

static int ex2(void)
{
  /* Local variables */
  Integer exit_status=0, i, j, n, tdh;
  double *g=0, *h=0, *h_central=0, *h_forward=0, *hess_diag=0, objf, *x=0;



  Nag_DerivInfo *deriv_info=0;
  Nag_E04_Opt options;
  NagError fail;

  INIT_FAIL(fail);


  n = MAXN;

  if (n>=1)
    {
      if ( !( x = NAG_ALLOC(n, double)) ||
           !( h_central = NAG_ALLOC(n, double)) ||
           !( h_forward = NAG_ALLOC(n, double)) ||
           !( g = NAG_ALLOC(n, double)) ||
           !( h = NAG_ALLOC(n*n, double)) ||
           !( hess_diag = NAG_ALLOC(n, double)) ||
           !( deriv_info = NAG_ALLOC(n, Nag_DerivInfo))
           )


        {
          Vprintf("Allocation failure\n");
          exit_status = -1;
          goto END;
        }
      tdh = n;
    }
  else
    {
      Vprintf("Invalid n.\n");
      exit_status = 1;
      return exit_status;
    }


  x[0] = 3.0;
  x[1] = -1.0;
  x[2] = 0.0;
  x[3] = 1.0;

  Vprintf("\nExample 2: some options are set\n");

  /* nag_opt_init (e04xxc).
   * Initialization function for option setting
   */
  nag_opt_init(&options);
  options.list = Nag_FALSE;
  options.print_deriv = Nag_D_NoPrint;
  options.deriv_want = Nag_Grad_HessDiag;

  Vprintf("\nEstimate gradient and Hessian diagonals given function only\n");

  /* Note: it is acceptable to pass an array of length n (hess_diag)
   * as the Hessian parameter in this case.
   */
  /* nag_opt_estimate_deriv (e04xac), see above. */
  nag_opt_estimate_deriv(n, x, objfun, &objf, g, h_forward, h_central,
                         hess_diag, tdh, deriv_info, &options, NAGCOMM_NULL,
                         &fail);
  if (fail.code != NE_NOERROR)
    {
      Vprintf("Error from nag_opt_estimate_deriv (e04xac).\n%s\n",
              fail.message);
      exit_status = 1;
      goto END;
    }

  Vprintf("\nFunction value: %12.4e\n", objf);
  Vprintf("Estimated gradient vector\n");
  for (i = 0; i < n; ++i)
    Vprintf("%12.4e  ", g[i]);
  Vprintf("\nEstimated Hessian matrix diagonal\n");
  for (i = 0; i < n; ++i)
    Vprintf("%12.4e  ", hess_diag[i]);
  Vprintf("\n");

  options.deriv_want = Nag_HessFull;

  Vprintf("\nEstimate full Hessian given function and gradients\n");
  /* nag_opt_estimate_deriv (e04xac), see above. */
  nag_opt_estimate_deriv(n, x, objfun, &objf, g, h_forward, h_central,
                         h, tdh, deriv_info, &options, NAGCOMM_NULL, &fail);
  if (fail.code != NE_NOERROR)
    {
      Vprintf("Error from nag_opt_estimate_deriv (e04xac).\n%s\n",
              fail.message);
      exit_status = 1;
      goto END;
    }


  Vprintf("\nFunction value: %12.4e\n", objf);
  Vprintf("Computed gradient vector\n");
  for (i = 0; i < n; ++i)
    Vprintf("%12.4e  ", g[i]);
  Vprintf("\nEstimated Hessian matrix\n");
  for (i = 0; i < n; ++i)
    {
      for (j = 0; j < n; ++j)
        Vprintf("%12.4e  ", H(i,j));
      Vprintf("\n");
    }
 END:
  if (x) NAG_FREE(x);
  if (h_central) NAG_FREE(h_central);
  if (h_forward) NAG_FREE(h_forward);
  if (g) NAG_FREE(g);
  if (h) NAG_FREE(h);
  if (hess_diag) NAG_FREE(hess_diag);
  if (deriv_info) NAG_FREE(deriv_info);
  return exit_status;

} /* ex2 */