//      g02gp Example Program Text
//      C# version, NAG Copyright 2008
using System;
using NagLibrary;
using System.Globalization;
using System.IO;
namespace NagDotNetExamples
{
  public class G02GPE
  {
    static string datafile = "ExampleData/g02gpe.d";
    static void Main(String[] args)
    {
      if (args.Length == 1)
      {
        datafile = args[0];
      }
      StartExample();
    }
    public static void StartExample()
    {
      try
      {
        DataReader sr = new DataReader(datafile);
        double a,   eps,   rss,   s,   tol; int i,  idf,  ip,  iprint,  irank,  j,  m,  maxit,  n;
        bool vfobs = false;
        string error = "",
        link = "",
        mean = "",
        offset = "",
        weight = "";
        int ifail;
        Console.WriteLine("g02gp Example Program Results");
        // 
        //      Skip headings in data file
        a = 0.0;
        sr.Reset();
        sr.Reset();
        //      Read in training data for model that will be used for prediction
        sr.Reset();
        link = sr.Next();
        mean = sr.Next();
        offset = sr.Next();
        weight = sr.Next();
        n = int.Parse(sr.Next());
        m = int.Parse(sr.Next());
        s = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
        iprint = int.Parse(sr.Next());
        double[] eta = new double[n];
        double[] fwt = new double[n];
        double[] off = new double[n];
        double[] pred = new double[n];
        double[] seeta = new double[n];
        double[] sepred = new double[n];
        double[] t = new double[1];
        double[] wt = new double[n];
        double[,] x = new double[n, m];
        double[] y = new double[n];
        int[] isx = new int[m];
        if (n < 2 || m < 1)
        {
          Console.Write("   ** Problem size is too small.");
          goto L200;
        }
        if ((weight == "W") || (weight == "w"))
        {
          for (i = 1; i <= n; i++)
          {
            sr.Reset();
            for (j = 1; j <= m; j++)
            {
              x[i - 1, j - 1] = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
            }
            y[i - 1] = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
            fwt[i - 1] = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
          }
        }
        else
        {
          for (i = 1; i <= n; i++)
          {
            sr.Reset();
            for (j = 1; j <= m; j++)
            {
              x[i - 1, j - 1] = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
            }
            y[i - 1] = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
          }
        }
        sr.Reset();
        for (j = 1; j <= m; j++)
        {
          isx[j - 1] = int.Parse(sr.Next());
        }
        //      Calculate ip
        ip = 0;
        for (j = 1; j <= m; j++)
        {
          if (isx[j - 1] > 0)
          {
            ip = ip + 1;
          }
        }
        if ((mean == "M") || (mean == "m"))
        {
          ip = ip + 1;
        }
        double[] b = new double[ip];
        double[] cov = new double[(ip*ip+ip)/2];
        double[,] v = new double[n, 7 + ip];
        double[] se = new double[ip];
        if ((link == "E") || (link == "e"))
        {
          sr.Reset();
          a = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
        }
        //      Set control parameters
        eps = 0.0000010e0;
        tol = 0.000050e0;
        maxit = 10;
        //      Call method to fit model to training data
        G02.g02ga(link, mean, offset, weight, n, x, m, isx, ip, y, fwt, ref s, a, out rss, out idf, b,
        out irank, se, cov, v, tol, maxit, iprint, eps, out ifail);
        if (ifail != 0)
        {
          Console.WriteLine("** g02ga failed with ifail = {0,5}", ifail);
          if (ifail < 6)
          {
            goto L200;
          }
        }
        //      Display parameter estimates for training data
        Console.WriteLine(" ");
        Console.WriteLine("  Residual sum of squares = {0,12:e4} Degrees of freedom = {1,2}", rss, idf);
        Console.WriteLine(" ");
        Console.Write(" {0}", "      Estimate     Standard error");
        Console.WriteLine(" ");
        for (i = 1; i <= ip; i++)
        {
          Console.WriteLine("  {0,14:f4}{1,14:f4}", b[i - 1], se[i - 1]);
        }
        // 
        //      Skip second lot of headings in data file
        sr.Reset();
        //      Read in data to predict from and check array sizes
        sr.Reset();
        n = int.Parse(sr.Next());
        vfobs = bool.Parse(sr.Next());
        offset = sr.Next();
        weight = sr.Next();
        if (n < 1)
        {
          Console.Write(" {0}", "n < 1");
          goto L200;
        }
        if ((offset == "Y") && (weight == "W"))
        {
          for (i = 1; i <= n; i++)
          {
            sr.Reset();
            for (j = 1; j <= m; j++)
            {
              x[i - 1, j - 1] = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
            }
            off[i - 1] = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
            wt[i - 1] = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
          }
        }
        else if (offset == "Y")
        {
          for (i = 1; i <= n; i++)
          {
            sr.Reset();
            for (j = 1; j <= m; j++)
            {
              x[i - 1, j - 1] = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
            }
            off[i - 1] = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
          }
        }
        else if (weight == "W")
        {
          for (i = 1; i <= n; i++)
          {
            sr.Reset();
            for (j = 1; j <= m; j++)
            {
              x[i - 1, j - 1] = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
            }
            wt[i - 1] = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
          }
        }
        else
        {
          for (i = 1; i <= n; i++)
          {
            sr.Reset();
            for (j = 1; j <= m; j++)
            {
              x[i - 1, j - 1] = double.Parse(sr.Next(), CultureInfo.InvariantCulture);
            }
          }
        }
        //      Using g02ga to fit training model, so error structure is normal
        error = "N";
        //      Call prediction method
        G02.g02gp(error, link, mean, offset, weight, n, x, m, isx, ip, t, off, wt, s, a, b, cov, vfobs,
        eta, seeta, pred, sepred, out ifail);
        // errfn, link, mean, offset, weight, n, x, m, isx, ip, t, off, wt, s, a, b, cov, vfobs, eta, seeta, pred, sepred, out ifail)
        if (ifail != 0)
        {
          Console.WriteLine("** g02gp failed with ifail = {0,5}", ifail);
          if (ifail != 22)
          {
            goto L200;
          }
        }
        //      Display predicted values
        Console.WriteLine(" ");
        Console.Write(" {0} {1}", "  I      ETA         SE(ETA)      Predicted  ",
        " SE(Predicted)");
        Console.WriteLine(" ");
        for (i = 1; i <= n; i++)
        {
          Console.WriteLine("  {0,3}){1,10:f5}        {2,10:f5}        {3,10:f5}        {4,10:f5}        ", i, eta[i - 1], seeta[i - 1], pred[i - 1], sepred[i - 1]);
        }
        // 
        L200: ;
        // 
      }
      catch (Exception e)
      {
        Console.WriteLine(e.Message);
        Console.Write( "Exception Raised");
      }
    }
  }
}