《概率统计与数据处理SAS 发酵猪饲料的应用研究》由会员分享，可在线阅读，更多相关《概率统计与数据处理SAS 发酵猪饲料的应用研究（4页珍藏版）》请在金锄头文库上搜索。
1、发酵猪饲料的应用研究饲料发酵后的酸度是评价饲料合格与否的一个重要指标，它受多个因素的影响，其中发酵温度、发酵时间、pH值和投曲度对其影响显著。本文利用16次发酵猪饲料试验数据和结果，采用逐步回归分析法选择适当的二次多项式，以预测酸度值。逐步回归分析法程序如下：(x1-x4分别代表发酵温度、发酵时间、pH值和投曲度，y代表饲料酸度)data e11;input x1-x4 y;cards;10 12 7 5 6.3610 24 6 10 7.4310 48 5 10 10.3 10 72 4 5 11.5620 12 6 10 8.6620 24 7 5 5.3920 48 4 5 15.520
2、 72 5 10 19.5330 12 5 5 12.0830 24 4 10 13.1330 48 7 10 8.0330 72 6 5 12.4550 12 4 10 13.4950 24 5 5 10.7750 48 6 5 9.850 72 7 10 16 64;proc print;run;proc reg data=e11;model y=x1-x4/selection=stepwise;run;运行结果如下：The SAS System 14:27 Saturday, December 17, 2008 1 Obs x1 x2 x3 x4 1 10 12 7 5 6.362 10
3、 24 6 10 7.433 10 48 5 10 10.304 10 72 4 5 11.565 20 12 6 10 8.666 20 24 7 5 5.397 20 48 4 5 15.508 20 72 5 10 19.539 30 12 5 5 12.0810 30 24 4 10 13.1311 30 48 7 10 8.0312 30 72 6 5 12.4513 50 12 4 10 13.4914 50 24 5 5 10.7715 50 48 6 5 9.8016 50 72 7 10 16.00The SAS System 14:27 Saturday, December
4、 17, 2008 2The REG ProcedureModel: MODEL1Dependent Variable: yStepwise Selection: Step 1Variable x2 Entered: R-Square = 0.2790 and C(p) = 12.8179Analysis of VarianceSum of MeanSource DF Squares Square F Value Pr FModel 1 59.46090 59.46090 5.42 0.0354Error 14 153.64110 10.97436Corrected Total 15 213.
5、10200Parameter StandardVariable Estimate Error Type II SS F Value Pr FIntercept 8.01733 1.62806 266.13177 24.25 0.0002x2 0.08366 0.03594 59.46090 5.42 0.0354Bounds on condition number: 1, 1-Stepwise Selection: Step 2Variable x3 Entered: R-Square = 0.5506 and C(p) = 5.4704The SAS System 14:27 Saturda
6、y, December 17, 2008 3The REG ProcedureModel: MODEL1Dependent Variable: yStepwise Selection: Step 2Analysis of VarianceSum of MeanSource DF Squares Square F Value Pr FModel 2 117.32892 58.66446 7.96 0.0055Error 13 95.77308 7.36716Corrected Total 15 213.10200Parameter StandardVariable Estimate Error
7、Type II SS F Value Pr FIntercept 17.37283 3.59474 172.07005 23.36 0.0003x2 0.08366 0.02945 59.46090 8.07 0.0139x3 -1.70100 0.60693 57.86802 7.85 0.0150Bounds on condition number: 1, 4-Stepwise Selection: Step 3Variable x1 Entered: R-Square = 0.6334 and C(p) = 4.6181Analysis of VarianceSum of MeanSou
8、rce DF Squares Square F Value Pr FModel 3 134.98662 44.99554 6.91 0.0059Error 12 78.11538 6.50962Corrected Total 15 213.10200The SAS System 14:27 Saturday, December 17, 2008 4The REG ProcedureModel: MODEL1Dependent Variable: yStepwise Selection: Step 3Parameter StandardVariable Estimate Error Type I
9、I SS F Value Pr FIntercept 15.41954 3.58114 120.68562 18.54 0.0010x1 0.07103 0.04313 17.65770 2.71 0.1255x2 0.08366 0.02768 59.46090 9.13 0.0106x3 -1.70100 0.57051 57.86802 8.89 0.0115Bounds on condition number: 1, 9-All variables left in the model are significant at the 0.1500 level.No other variab
10、le met the 0.1500 significance level for entry into the model.Summary of Stepwise SelectionVariable Variable Number Partial Model Entered Removed Vars In R-Square R-Square C(p) F value PrFx2 1 0.2790 0.2790 12.8179 5.42 0.0354x3 2 0.2716 0.5506 5.4704 7.85 0.0150x1 3 0.0829 0.6334 4.6181 2.71 0.1255在0.15水平下得到的回归方程为y=15.41954+0.07104x1+0.08366x2-1.7.100x3,此时Pr为0.0059，模型高度显著。
