using Accord.MachineLearning;

using ExcelDataReader;

using System;

using System.Collections.Generic;

using System.ComponentModel;

using System.Data;

using System.Drawing;

using System.IO;

using System.Linq;

using System.Text;

using System.Threading.Tasks;

using System.Windows.Forms;

​

namespace MachineLearning

{

    public partial class Form1 : Form

    {

        private string filePath = @"C:\Users\Duy Nguyen\Downloads\DTB_IT.xlsx";

​

        public Form1()

        {

            InitializeComponent();

        }

​

        private void button1_Click(object sender, EventArgs e)

        {

            FileStream stream = File.Open(filePath, FileMode.Open, FileAccess.Read);

            IExcelDataReader excelReader;

​

            //1. Reading Excel file

            if (Path.GetExtension(filePath).ToUpper() == ".XLS")

            {

                //1.1 Reading from a binary Excel file ('97-2003 format; *.xls)

                excelReader = ExcelReaderFactory.CreateBinaryReader(stream);

            }

            else

            {

                //1.2 Reading from a OpenXml Excel file (2007 format; *.xlsx)

                excelReader = ExcelReaderFactory.CreateOpenXmlReader(stream);

            }

            //2. DataSet - The result of each spreadsheet will be created in the result.Tables

            DataSet result = excelReader.AsDataSet(new ExcelDataSetConfiguration()

            {

                // Gets or sets a value indicating whether to set the DataColumn.DataType 

                // property in a second pass.

                UseColumnDataType = true,

​

                // Gets or sets a callback to determine whether to include the current sheet

                // in the DataSet. Called once per sheet before ConfigureDataTable.

                FilterSheet = (tableReader, sheetIndex) => true,

​

                // Gets or sets a callback to obtain configuration options for a DataTable. 

                ConfigureDataTable = (tableReader) => new ExcelDataTableConfiguration()

                {

                    // Gets or sets a value indicating the prefix of generated column names.

                    EmptyColumnNamePrefix = "Column",

​

                    // Gets or sets a value indicating whether to use a row from the 

                    // data as column names.

                    UseHeaderRow = true,

​

                }

            });

            DataTable dt = result.Tables[0];

​

            List<double[]> data = new List<double[]>();

​

            foreach (DataRow dtRow in dt.Rows)

            {

                List<double> row = new List<double>();

                for(int i = 4; i <= 12; i++)

                {

                    double val = Double.Parse(dtRow[i].ToString());

                    row.Add(val);

                }

                data.Add(row.ToArray());

            }

            double[][] input = data.ToArray();

​

​

​

            KMeans kmeans = new KMeans(k: 2);

            //Hoc huan luyen

            var clusters = kmeans.Learn(input);

            //

            int[] labels = clusters.Decide(input);

​

            dt.Columns.Add("Cluster", typeof(System.Int32));

            foreach (DataRow dtRow in dt.Rows)

            {

                int index = dt.Rows.IndexOf(dtRow);

                dtRow["Cluster"] = labels[index];

            }

​

            dataGridView1.DataSource = dt;

​

            DataTable dt_lb1 = dt.AsEnumerable().Where(row => row.Field<Int32>("Cluster") == 0).CopyToDataTable();

            DataTable dt_lb2 = dt.AsEnumerable().Where(row => row.Field<Int32>("Cluster") == 1).CopyToDataTable();

​

            dataGridView2.DataSource = dt_lb1;

            dataGridView3.DataSource = dt_lb2;

        }

    }

}