fitlm函数产生的参数问题
1.p值是每个自变量在方程中所占的重要性吗?换言之,可以根据p值大小比较各自变量对方程的影响程度吗?
2.F-statistic vs constant model: 4.35,这是个什么意思
可能问题比较白痴,但是真的是急求有人解答,跪谢,附上我的运算程序
>> clear
x1=[34 28.001 28 28 28 22.001 22 22 22 22 16.001 16 16 16 10.001 10 10 10 10 4.001 4 4 4 4];
x2=[12 6 6 6.001 6 1 1 1.001 1 1 15 15 15.001 15 9 9 9.001 9 9 3 3 3.001 3 3];
x3=[8 2 2 2 2.001 10 10 10 10.001 10 4 4 4 4.001 12 12 12 12.001 12 6 6 6 6.001 6];
y=[9.78 14.12 14.34 9.08 12.49 9.84 12.86 11.36 14.96 12.33 7.19 10.29 12.64 9.45 16.53 13.64 15.58 15.71 15.27 13.58 17.71 14.00 9.51 14.72]';
x=[x1',x2',x3'];
fitlm(x, y,'quadratic');
>>
>> ans
ans =
Linear regression model:
y ~ 1 + x1*x2 + x1*x3 + x2*x3 + x1^2 + x2^2 + x3^2
Estimated Coefficients:
Estimate SE tStat pValue
________ ______ _______ _________
(Intercept) 60641 18548 3.2694 0.0055927
x1 -1509.4 847.11 -1.7818 0.096477
x2 -3103.3 2782.3 -1.1153 0.28348
x3 -11908 3742.8 -3.1816 0.0066603
x1:x2 -73.932 64.765 -1.1415 0.2728
x1:x3 193.36 76.495 2.5277 0.024137
x2:x3 336.27 153.26 2.1941 0.045605
x1^2 19.076 30.905 0.61723 0.54699
x2^2 139.7 91.156 1.5325 0.14768
x3^2 411.14 141.57 2.904 0.01155
Number of observations: 24, Error degrees of freedom: 14
Root Mean Squared Error: 1.79
R-squared: 0.737, Adjusted R-Squared 0.567
F-statistic vs. constant model: 4.35, p-value = 0.00719