private void btnRegress_Click(object sender, EventArgs e)
{
if (pls == null || dgvRegressionInput.DataSource == null)
{
MessageBox.Show("Please compute the analysis first.");
return;
}
int components = (int)numComponentsRegression.Value;
regression = pls.CreateRegression(components);
DataTable table = dgvRegressionInput.DataSource as DataTable;
DataTable inputTable = table.DefaultView.ToTable(false, inputColumnNames);
DataTable outputTable = table.DefaultView.ToTable(false, outputColumnNames);
double[][] sourceInput = Matrix.ToArray(inputTable as DataTable);
double[][] sourceOutput = Matrix.ToArray(outputTable as DataTable);
double[,] result = Matrix.ToMatrix(regression.Compute(sourceInput));
double[] rSquared = regression.CoefficientOfDetermination(sourceInput, sourceOutput, cbAdjusted.Checked);
dgvRegressionOutput.DataSource = new ArrayDataView(result, outputColumnNames);
dgvRSquared.DataSource = new ArrayDataView(rSquared, outputColumnNames);
}
//EXAMPLE: http://accord-framework.net/docs/html/T_Accord_Statistics_Models_Regression_Linear_MultivariateLinearRegression.htm
static void Main(string[] args)
{
CSV_Parser parser = new CSV_Parser();
RegressionData data = parser.ParseDataFile();
// Use Ordinary Least Squares to create the regression
OrdinaryLeastSquares ols = new OrdinaryLeastSquares();
// Now, compute the multivariate linear regression:
MultivariateLinearRegression regression = ols.Learn(data.InterestRatings, data.MajorRatings);
// We can obtain predictions using
double[][] predictions = regression.Transform(data.InterestRatings);
// The prediction error is
double error = new SquareLoss(data.MajorRatings).Loss(predictions); // 0
// We can also check the r-squared coefficients of determination:
//double[] r2 = regression.CoefficientOfDetermination(topicRatings, majorRatings);
double[][] r2 = regression.Weights;
Console.WriteLine("WEIGHTS:");
//writeCSVfile(data, r2);
GenerateCSFile(data, r2);
Console.WriteLine("Coefficient Of Determination");
double[] r3 = regression.CoefficientOfDetermination(data.InterestRatings, data.MajorRatings);
for (int i = 0; i < r3.Length; i++)
{
Console.WriteLine(r3[i]);
}
Console.Read();
}
public void RegressTest2()
{
// The multivariate linear regression is a generalization of
// the multiple linear regression. In the multivariate linear
// regression, not only the input variables are multivariate,
// but also are the output dependent variables.
// In the following example, we will perform a regression of
// a 2-dimensional output variable over a 3-dimensional input
// variable.
double[][] inputs =
{
// variables: x1 x2 x3
new double[] { 1, 1, 1 }, // input sample 1
new double[] { 2, 1, 1 }, // input sample 2
new double[] { 3, 1, 1 }, // input sample 3
};
double[][] outputs =
{
// variables: y1 y2
new double[] { 2, 3 }, // corresponding output to sample 1
new double[] { 4, 6 }, // corresponding output to sample 2
new double[] { 6, 9 }, // corresponding output to sample 3
};
// With a quick eye inspection, it is possible to see that
// the first output variable y1 is always the double of the
// first input variable. The second output variable y2 is
// always the triple of the first input variable. The other
// input variables are unused. Nevertheless, we will fit a
// multivariate regression model and confirm the validity
// of our impressions:
// Create a new multivariate linear regression with 3 inputs and 2 outputs
var regression = new MultivariateLinearRegression(3, 2);
// Now, compute the multivariate linear regression:
double error = regression.Regress(inputs, outputs);
// At this point, the regression error will be 0 (the fit was
// perfect). The regression coefficients for the first input
// and first output variables will be 2. The coefficient for
// the first input and second output variables will be 3. All
// others will be 0.
//
// regression.Coefficients should be the matrix given by
//
// double[,] coefficients = {
// { 2, 3 },
// { 0, 0 },
// { 0, 0 },
// };
//
// The first input variable coefficients will be 2 and 3:
Assert.AreEqual(2, regression.Coefficients[0, 0], 1e-10);
Assert.AreEqual(3, regression.Coefficients[0, 1], 1e-10);
// And all other coefficients will be 0:
Assert.AreEqual(0, regression.Coefficients[1, 0], 1e-10);
Assert.AreEqual(0, regression.Coefficients[1, 1], 1e-10);
Assert.AreEqual(0, regression.Coefficients[2, 0], 1e-10);
Assert.AreEqual(0, regression.Coefficients[2, 1], 1e-10);
// We can also check the r-squared coefficients of determination:
double[] r2 = regression.CoefficientOfDetermination(inputs, outputs);
// Which should be one for both output variables:
Assert.AreEqual(1, r2[0]);
Assert.AreEqual(1, r2[1]);
foreach (var e in regression.Coefficients)
{
Assert.IsFalse(double.IsNaN(e));
}
Assert.AreEqual(0, error, 1e-10);
Assert.IsFalse(double.IsNaN(error));
}
private double[][] CreateWeights(List <Stock> listOfStocks)
{
// The multivariate linear regression is a generalization of
// the multiple linear regression. In the multivariate linear
// regression, not only the input variables are multivariate,
// but also are the output dependent variables.
// In the following example, we will perform a regression of
// a 2-dimensional output variable over a 3-dimensional input
// variable.
var inputs = new double[listOfStocks.Count][];
var outputs = new double[listOfStocks.Count][];
for (int i = 0; i < listOfStocks.Count; i++)
{
inputs[i] = new[]
{
listOfStocks[i].AdjClose,
listOfStocks[i].Close,
listOfStocks[i].High,
listOfStocks[i].Low,
listOfStocks[i].Open,
};
outputs[i] = new[] { listOfStocks[i].Volume };
}
// With a quick eye inspection, it is possible to see that
// the first output variable y1 is always the double of the
// first input variable. The second output variable y2 is
// always the triple of the first input variable. The other
// input variables are unused. Nevertheless, we will fit a
// multivariate regression model and confirm the validity
// of our impressions:
// Use Ordinary Least Squares to create the regression
OrdinaryLeastSquares ols = new OrdinaryLeastSquares();
// Now, compute the multivariate linear regression:
MultivariateLinearRegression regression = ols.Learn(inputs, outputs);
// We can obtain predictions using
double[][] predictions = regression.Transform(inputs);
// The prediction error is
double error = new SquareLoss(outputs).Loss(predictions); // 0
// At this point, the regression error will be 0 (the fit was
// perfect). The regression coefficients for the first input
// and first output variables will be 2. The coefficient for
// the first input and second output variables will be 3. All
// others will be 0.
//
// regression.Coefficients should be the matrix given by
//
// double[,] coefficients = {
// { 2, 3 },
// { 0, 0 },
// { 0, 0 },
// };
//
// We can also check the r-squared coefficients of determination:
double[] r2 = regression.CoefficientOfDetermination(inputs, outputs);
return(regression.Weights);
}
/// <summary>
/// 求出输入数据的一次模型 和 二次模型 输出顺序为 一次斜率k 截距b 模型拟合程度r 二次参数a 一次参数b 截距c 模型拟合程度r
/// </summary>
/// <param name="input"></param>
/// <returns></returns>
public static List <PValue> getModel(List <PValue> input)
{
double MTAEquRLK = 0; //k
double MTAEquRLB = 0; //b
double MTAEquRLR = 0; //r
double MTAEquRQA = 0; //a
double MTAEquRQB = 0; //b
double MTAEquRQC = 0; //c
double MTAEquRQR = 0; //r
int length = input.Count;
double[] X = new double[length];
double[] Y = new double[length];
for (int i = 0; i < length; i++)
{
X[i] = i + 1;
Y[i] = input[i].Value;
}
OrdinaryLeastSquares ols = new OrdinaryLeastSquares();
SimpleLinearRegression regression = ols.Learn(X, Y);
MTAEquRLK = regression.Slope; //一次斜率
MTAEquRLB = regression.Intercept; //截距
MTAEquRLR = regression.CoefficientOfDetermination(X, Y); //R^2代表模型拟合程度
double[][] MultiX = new double[X.Length][];
for (int i = 0; i < X.Length; i++)
{
double x = X[i];
MultiX[i] = new double[] { x *x, x };
}
double[][] MultiY = new double[X.Length][];
//MultiY =X'
for (int i = 0; i < X.Length; i++)
{
double y = Y[i];
MultiY[i] = new double[] { y };
}
OrdinaryLeastSquares olsMulti = new OrdinaryLeastSquares();
MultivariateLinearRegression regressionMulti = olsMulti.Learn(MultiX, MultiY);
double[][] weights = regressionMulti.Weights;
MTAEquRQA = weights[0][0]; //二次参数
MTAEquRQB = weights[1][0]; //一次参数
MTAEquRQC = regressionMulti.Intercepts[0]; //截距
MTAEquRQR = regressionMulti.CoefficientOfDetermination(MultiX, MultiY)[0]; //R^2代表模型拟合程度
PValue LK = new PValue(); LK.Value = MTAEquRLK;
PValue LB = new PValue(); LB.Value = MTAEquRLB;
PValue LR = new PValue(); LR.Value = MTAEquRLR;
PValue QA = new PValue(); QA.Value = MTAEquRQA;
PValue QB = new PValue(); QB.Value = MTAEquRQB;
PValue QC = new PValue(); QC.Value = MTAEquRQC;
PValue QR = new PValue(); QR.Value = MTAEquRQR;
List <PValue> list = new List <PValue>();
list.Add(LK);
list.Add(LB);
list.Add(LR);
list.Add(QA);
list.Add(QB);
list.Add(QC);
list.Add(QR);
return(list);
}
public static Results Calcu(List <PValue>[] inputs, CalcuInfo calcuinfo)
{
//公用变量
bool _errorFlag = false;
string _errorInfo = "";
bool _warningFlag = false;
string _warningInfo = "";
bool _fatalFlag = false;
string _fatalInfo = "";
int i;
//0输出初始化：该算法如果没有有效输入值（inputs为null）或者输入值得有效值为null，则应当有输出，大部分的输出项为0。个别输出项为空
List <PValue>[] results = new List <PValue> [7];
for (i = 0; i < results.Length; i++)
{
results[i] = new List <PValue>();
results[i].Add(new PValue(0, calcuinfo.fstarttime, calcuinfo.fendtime, (long)StatusConst.InputIsNull));
}
try
{
//数据量分析
//——MMTAAnaBase对偏差的计算结果，有19个值，一个设备100个温度点，那么也就1900个点。readmulti读取速度不会太慢。
//——本计算仅用到19个计算结果中的均值，但是为了配置方便，仍然按MMTAAnaBase计算结果全部取得方式
int MMTAAnaBaseOutputNumber = 19; //本算法是直接读取MMTAAnaBase算法所有计算结果。MMTAAnaBase算法有19个计算结果。要靠次参数完成对计算结果的解析。
int validDataNumber = 0;
//0、输入：输入是同一设备上所有温度点的小时均值，或者天均值
//0.1、输入处理：输入长度。当输入为空时，则输出项也为空.
if (inputs == null)
{
return(new Results(results, _errorFlag, _errorInfo, _warningFlag, _warningInfo, _fatalFlag, _fatalInfo)); //不报错，直接返回默认值。整体为空计算引擎已经记录报警或错误。
}
//0.2、输入处理：截止时刻值。该算法不需要截止时刻点参与计算。
//输入的是该设备上所有温度测点每小时基本统计值，每个统计值也就是每个测点，仅有两个值，一个是小时值，一个是截止时刻值
string[] sourtagname = calcuinfo.sourcetagname.Split(';'); //本算法是做管子位置与温度的回归。管子的位置需要从标签名称中取
List <List <PValue> > AvgInputsList = new List <List <PValue> >(); //将MMTAAnaBase算法结果中的均值取出来
List <string> AvgtagnameList = new List <string>(); //将均值对应的标签名称取出来
for (i = 0; i < inputs.Length; i++)
{
//按MMTAAnaBase算法,将均值（MMTAAnaBaseOutputNumber==1 ）取出
//对于设备来说，某些单点状态位异常，直接过滤而不进行报警。只有这一类计算的结果全部异常才报警。
if (i % MMTAAnaBaseOutputNumber == 1 && inputs[i] != null && inputs[i].Count >= 2 && inputs[i][0].Status == 0)
{
//找到所有有效均值的标签名
AvgtagnameList.Add(sourtagname[i]);
//找到数据
//if (AvgInputsList[i / MMTAAnaBaseOutputNumber] == null)
AvgInputsList.Add(new List <PValue>());
//只取第一个值
AvgInputsList[validDataNumber].Add(new PValue(inputs[i][0].Value, inputs[i][0].Timestamp, inputs[i][0].Endtime, inputs[i][0].Status));
validDataNumber = validDataNumber + 1;
}
}//end for
string[] Avgtagname = AvgtagnameList.ToArray(); //有效点不同，标签数量不同，删除后面的空位。
List <PValue>[] AvgInputs = AvgInputsList.ToArray(); //与标签点对应添加的有效数据
if (validDataNumber < 2)
{
_warningFlag = true;
_warningInfo = "当前时间段过滤非正常状态点后，全部点的均值有效个数少于2个，无法进行回归计算！";
return(new Results(results, _errorFlag, _errorInfo, _warningFlag, _warningInfo, _fatalFlag, _fatalInfo));
}
//定义变量
double MTAEquRLK = 0;
double MTAEquRLB = 0;
double MTAEquRLR = 0;
double MTAEquRQA = 0;
double MTAEquRQB = 0;
double MTAEquRQC = 0;
double MTAEquRQR = 0;
//计算
//1、线性回归
//参考 http://accord-framework.net/docs/html/T_Accord_Statistics_Models_Regression_Linear_SimpleLinearRegression.htm
List <double> ArrayX = new List <double>(); //回归分析，自变量是管号
List <double> ArrayY = new List <double>(); //回归分析，应变量是温度
for (i = 0; i < AvgInputs.Length; i++)
{
double xposition = double.Parse(Avgtagname[i].Substring(7, 2));
for (int j = 0; j < AvgInputs[i].Count; j++)
{
ArrayX.Add(xposition);
ArrayY.Add(AvgInputs[i][j].Value);
}
}
/*
* //测试数据
* ArrayX = new List<double> { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 };
* ArrayY = new List<double> { 2, 4, 6, 8, 10, 12, 14, 16, 18, 20 };
* //按上述测试数据，MTAEquRLK=2，MTAEquRLB=0，MTAEquRLRR=1
*
*
* ArrayX = new double[10] { 1.2, 1.9, 3.1, 3.7, 5, 6, 7.1, 7.9, 9, 10 };
* ArrayY = new double[10] { 2.1, 4.2, 6.8, 8.5, 10, 12, 14.6, 14, 18, 20 };
*/
//按上述测试数据，MTAEquRLK=1.911，MTAEquRLB=0.526，MTAEquRLRR=0.985
double[] X = ArrayX.ToArray();
double[] Y = ArrayY.ToArray();;
OrdinaryLeastSquares ols = new OrdinaryLeastSquares();
SimpleLinearRegression regression = ols.Learn(X, Y);
MTAEquRLK = regression.Slope; //一次斜率
MTAEquRLB = regression.Intercept; //截距
MTAEquRLR = regression.CoefficientOfDetermination(X, Y); //R^2代表模型拟合程度
//2、多项式回归
//参考 http://accord-framework.net/docs/html/T_Accord_Statistics_Models_Regression_Linear_MultivariateLinearRegression.htm
//MultiX =[X^2 ,X]
double[][] MultiX = new double[X.Length][];
for (i = 0; i < X.Length; i++)
{
double x = X[i];
MultiX[i] = new double[] { x *x, x };
}
double[][] MultiY = new double[X.Length][];
//MultiY =X'
for (i = 0; i < X.Length; i++)
{
double y = Y[i];
MultiY[i] = new double[] { y };
}
//测试数据
/*
* MultiX = new double[5][];
* MultiX[0] = new double[2] { 1, 1};
* MultiX[1] = new double[2] { 4, 2};
* MultiX[2] = new double[2] { 9, 3};
* MultiX[3] = new double[2] { 16, 4};
* MultiX[4] = new double[2] { 25, 5};
*
* MultiY = new double[5][];
* MultiY[0] = new double[1] { 7 };
* MultiY[1] = new double[1] { 17 };
* MultiY[2] = new double[1] { 31 };
* MultiY[3] = new double[1] { 49 };
* MultiY[4] = new double[1] { 71 };
*/
//按上述测试数据，MTAEquRQA=2，MTAEquRQB=4，MTAEquRQC=1，MTAEquRQR=1
/*
* MultiX = new double[5][];
* MultiX[0] = new double[2] { 1, 1 };
* MultiX[1] = new double[2] { 4, 2 };
* MultiX[2] = new double[2] { 9, 3 };
* MultiX[3] = new double[2] { 16, 4 };
* MultiX[4] = new double[2] { 25, 5 };
*
* MultiY = new double[5][];
* MultiY[0] = new double[1] { 7.1 };
* MultiY[1] = new double[1] { 16.9 };
* MultiY[2] = new double[1] { 32 };
* MultiY[3] = new double[1] { 51 };
* MultiY[4] = new double[1] { 69 };
*/
//按上述测试数据，MTAEquRQA=1.45，MTAEquRQB=7.08，MTAEquRQC=-2，MTAEquRQR=0.998
OrdinaryLeastSquares olsMulti = new OrdinaryLeastSquares();
MultivariateLinearRegression regressionMulti = olsMulti.Learn(MultiX, MultiY);
double[][] weights = regressionMulti.Weights;
MTAEquRQA = weights[0][0]; //二次参数
MTAEquRQB = weights[1][0]; //一次参数
MTAEquRQC = regressionMulti.Intercepts[0]; //截距
MTAEquRQR = regressionMulti.CoefficientOfDetermination(MultiX, MultiY)[0]; //R^2代表模型拟合程度
//组织计算结果
results[0] = new List <PValue>();
results[0].Add(new PValue(MTAEquRLK, calcuinfo.fstarttime, calcuinfo.fendtime, 0));
results[1] = new List <PValue>();
results[1].Add(new PValue(MTAEquRLB, calcuinfo.fstarttime, calcuinfo.fendtime, 0));
results[2] = new List <PValue>();
results[2].Add(new PValue(MTAEquRLR, calcuinfo.fstarttime, calcuinfo.fendtime, 0));
results[3] = new List <PValue>();
results[3].Add(new PValue(MTAEquRQA, calcuinfo.fstarttime, calcuinfo.fendtime, 0));
results[4] = new List <PValue>();
results[4].Add(new PValue(MTAEquRQB, calcuinfo.fstarttime, calcuinfo.fendtime, 0));
results[5] = new List <PValue>();
results[5].Add(new PValue(MTAEquRQC, calcuinfo.fstarttime, calcuinfo.fendtime, 0));
results[6] = new List <PValue>();
results[6].Add(new PValue(MTAEquRQR, calcuinfo.fstarttime, calcuinfo.fendtime, 0));
return(new Results(results, _errorFlag, _errorInfo, _warningFlag, _warningInfo, _fatalFlag, _fatalInfo));
}
catch (Exception ex)
{
//计算中出任何错误，则需要记录log
//LogHelper.Write(LogType.Error, "计算模块错误!");
//记录计算模块的名称、当前标签、起始时间、结束时间
//string moduleInfo = string.Format("——计算模块的名称是：{0}，当前计算源标签是：{1}，计算起始时间是：{2}，计算结束时间是：{3}。", calcuinfo.fmodulename, calcuInfo.sourcetagname, calcuinfo.fstarttime.ToString(), calcuinfo.fendtime.ToString());
//LogHelper.Write(LogType.Error, moduleInfo);
//计算引擎报错具体信息
//string errInfo = string.Format("——具体报错信息：{0}。", ex.ToString());
//LogHelper.Write(LogType.Error, errInfo);
//返回null供计算引擎处理
_fatalFlag = true;
_fatalInfo = ex.ToString();
return(new Results(results, _errorFlag, _errorInfo, _warningFlag, _warningInfo, _fatalFlag, _fatalInfo));
}
}
