试验设计和数据处理办法回归分析.ppt

* * * * * * * * * * * * * * * * 一、SAS/INSIGHT 步骤：Solutions→Analysis→Interactive data analysis (在SAS/INSIGHT：OPEN窗 口中) → work →选中数据名 (本例为Data corn见V8 文件）→(OPEN)→Analyze →Fit [在Fit窗口中同时选中两个自变量(本例为x、 y)然后送入EXPAND中, x、y、x*x、x*y、 y*y即刻出现在X中,将因变量(本例为h)送入 Y中] →(OK) [若多项式超过2次, 在此窗口中点击INTERCEPT2 右侧的方块, 选择所需次数.] 二、SAS编程 data corn; do x=2 to 3 by 0.25 ; /*数据h可看成带行、列下标的变量*/ do y = 6 to 4 by -0.5 ; input h @@; x2=x*x; xy=x*y; y2=y*y; output; end; end; cards; -11 0 10 20 31 -11 -1 9 19 30 -12 -2 8 19 29 -13 -2 7 18 28 -14 13 6 17 27 ; proc reg data=corn; model h=t1-t5; run; The REG Procedure Model: MODEL1 Dependent Variable: h Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr F Model 5 4970.41143 994.08229 88.43 .0001 Error 19 213.58857 11.24150 Corrected Total 24 5184.00000 Root MSE 3.35284 R-Square 0.9588 Dependent Mean 9.00000 Adj R-Sq 0.9480 Coeff Var 37.25372 Parameter Estimates Parameter Standard Variable DF Estimate Error t Value Pr |t| Intercept 1 170.20000 65.22449 2.61 0.0172 t1 1 -49.48571 34.80310 -1.42 0.1713 t2 1 -19.65143 17.40155 -1.13 0.2728 t3 1 6.857