1

我正在尝试使用 org.apache.commons.math3 OLSMultipleLinearRegression.java：

OLSMultipleLinearRegression regression = new OLSMultipleLinearRegression();

我的 x 和 y 的数据格式在每次迭代中都不同，如下所示：

第一次迭代：

double[] y = {0.2342342};
double[][]x = { {0.234234234, 0.234165745, 0.234165745}};

第二次迭代

double[] y = {0.2342342, 0.23423423};
double[][]x = {
    {0.234234234, 0.234165745, 0.234165745},
    {0.234165745, 0.234165745, 0.234165745}
};

第三次迭代

double[] y = {0.2342342, 0.23423423, 0.234234234};
double[][]x = { 
    {0.234234234, 0.234165745, 0.234165745},
    {0.234234234, 0.234165745, 0.234165745},
    {0.234234234, 0.234165745, 0.234165745}
    };

第四次迭代 

double[] y = {0.2342342, 0.23423423, 0.234234234, 0.242312312};
double[][]x = { 
    {0.234234234, 0.234165745, 0.234165745},
    {0.234234234, 0.234165745, 0.234165745},
    {0.234234234, 0.234165745, 0.234165745},
    {0.234234234, 0.234165745, 0.234165745} 
    };

这是我的代码：

try {
     regression.newSampleData(y, x);
} catch (IllegalArgumentException e) {
     return getFallbackVmAllocationPolicy().isHostOverUtilized(host);
}

double[] estimates = regression.estimateRegressionParameters();

double predictedUtilization = 0.0;

for (int i = 0; i < utilizationHistory.length; i++) {
      predictedUtilization = estimates[0] + (estimates[1] * utilizationHistory[i])
                            + (estimates[2] * ramUtilizationHistory[i]) + (estimates[3] * bwUtilizationHistory[i]);
}

//predictedUtilization *= getSafetyParameter();

addHistoryEntry(host, predictedUtilization);
return predictedUtilization >= 1;

在执行此操作时，我得到一个异常，如下所示：

org.apache.commons.math3.linear.SingularMatrixException: matrix is singular
at org.apache.commons.math3.linear.QRDecomposition$Solver.solve(QRDecomposition.java:347)
at org.apache.commons.math3.stat.regression.OLSMultipleLinearRegression.calculateBeta(OLSMultipleLinearRegression.java:228)

我该如何解决这个异常？

4

0 回答 0