|
|
楼主
楼主 |
发表于 2006-2-16 14:58:41
|
只看该作者
使用sas处理二次螺旋回归错误出在何处？
我是一个新手，想使用sas处理二次螺旋回归试验结果，求出x1、x2、x3、x4对y的影响，我按照一本书上的方法进行设计，可是我的结果很不好，麻烦各位老师帮忙，小弟不胜感激！
我的数据：
x1 x2 x3 x4 y
0.2 0.2 0.2 0.2 9221.998
0.2 0.2 0.4 0.4 11838.04425
0.2 0.4 0.2 0.4 21585.135
0.2 0.4 0.4 0.2 24606.295
0.4 0.2 0.2 0.4 19372.00467
0.4 0.2 0.4 0.2 18111.1345
0.4 0.4 0.2 0.2 22884.45333
0.4 0.4 0.4 0.4 21162.832
0.01 0.3 0.3 0.3 14718.52933
0.59 0.3 0.3 0.3 18906.81233
0.3 0.01 0.3 0.3 11297.71775
0.3 0.59 0.3 0.3 16566.79567
0.3 0.3 0.01 0.3 20985.603
0.3 0.3 0.59 0.3 17151.01
0.3 0.3 0.3 0.01 25149.89617
0.3 0.3 0.3 0.59 23320.57067
0.3 0.3 0.3 0.3 12817.5085
0.3 0.3 0.3 0.3 29659.92967
0.3 0.3 0.3 0.3 21238.71908
我的结果（采用RSREG处理）：
The SAS System 11:39 Friday, February 16, 2006 18
The RSREG Procedure
Coding Coefficients for the Independent Variables
Factor Subtracted off Divided by
x1 0.300000 0.290000
x2 0.300000 0.290000
x3 0.300000 0.290000
x4 0.300000 0.290000
Response Surface for Variable y
Response Mean 18979
Root MSE 5496.224828
R-Square 0.5854
Coefficient of Variation 28.9600
Type I Sum
Regression DF of Squares R-Square F Value Pr > F
Linear 4 121464859 0.2382 1.01 0.4650
Quadratic 4 124702617 0.2445 1.03 0.4537
Crossproduct 3 52380615 0.1027 0.58 0.6477
Total Model 11 298548091 0.5854 0.90 0.5809
Sum of
Residual DF Squares Mean Square F Value Pr > F
Lack of Fit 5 69625856 13925171 0.20 0.9378
Pure Error 2 141833556 70916778
Total Error 7 211459411 30208487
The SAS System 11:39 Friday, February 16, 2006 19
The RSREG Procedure
Parameter
Estimate
Standard from Coded
Parameter DF Estimate Error t Value Pr > |t| Data
Intercept 1 -26433 33562 -0.79 0.4568 20279
x1 1 139659 90091 1.55 0.1650 3087.523441
x2 1 136374 90091 1.51 0.1739 5488.721104
x3 1 40090 90091 0.44 0.6698 -989.131463
x4 1 -28380 36325 -0.78 0.4602 -721.018868
x1*x1 1 -45102 57682 -0.78 0.4599 -3793.094712
x2*x1 1 -232091 194321 -1.19 0.2712 -19519
x2*x2 1 -79352 57682 -1.38 0.2113 -6673.504712
x3*x1 1 -107746 194321 -0.55 0.5965 -9061.449113
x3*x2 1 -695.375000 194321 -0.00 0.9972 -58.481037
x3*x3 1 -18281 57682 -0.32 0.7605 -1537.459712
x4*x1 0 0 . . . 0
x4*x2 0 0 . . . 0
x4*x3 0 0 . . . 0
x4*x4 1 43157 57682 0.75 0.4787 3629.470288
Sum of
Factor DF Squares Mean Square F Value Pr > F
x1 4 98982685 24745671 0.82 0.5519
x2 4 189171519 47292880 1.57 0.2834
x3 4 15209515 3802379 0.13 0.9684
x4 2 18444027 9222013 0.31 0.7463 |
|
