统计建模与R软件第六章课后习题答案.docx

1、统计建模与R软件第六章课后习题答案统计建模与R软件第六章习题答案(回归分析)Ex6.1（1） x y plot(x,y)由此判断，Y和X有线性关系。（2） lm.sol summary(lm.sol)Call:lm(formula = y 1 + x)Residuals: Min 1Q Median 3Q Max-128.591 -70.978 -3.727 49.263 167.228Coefficients: Estimate Std. Error t value Pr(|t|)(Intercept) 140.95 125.11 1.127 0.293x 364.18 19.26 18.9

2、08 6.33e-08 *-Signif. codes: 0 * 0.001 * 0.01 * 0.05 . 0.1 1Residual standard error: 96.42 on 8 degrees of freedomMultiple R-squared: 0.9781, Adjusted R-squared: 0.9754F-statistic: 357.5 on 1 and 8 DF, p-value: 6.33e-08回归方程为 Y=140.95+364.18X（3）1项很显著，但常数项0不显著。回归方程很显著。（4） new lm.pred lm.pred fit lwr u

3、pr1 2690.227 2454.971 2925.484故Y(7)= 2690.227, 2454.971,2925.484Ex6.2(1)pho-data.frame(x1 - c(0.4,0.4,3.1,0.6,4.7,1.7,9.4,10.1,11.6,12.6,10.9,23.1,23.1,21.6,23.1,1.9,26.8,29.9), x2 - c(52,34,19,34,24,65,44,31,29,58,37,46,50,44,56,36,58,51), x3 - c(158,163,37,157,59,123,46,117,173,112,111,114,134,73,

4、168,143,202,124), y lm.sol summary(lm.sol)Call:lm(formula = y x1 + x2 + x3, data = pho)Residuals: Min 1Q Median 3Q Max-27.575 -11.160 -2.799 11.574 48.808Coefficients: Estimate Std. Error t value Pr(|t|)(Intercept) 44.9290 18.3408 2.450 0.02806 *x1 1.8033 0.5290 3.409 0.00424 *x2 -0.1337 0.4440 -0.3

5、01 0.76771x3 0.1668 0.1141 1.462 0.16573-Signif. codes: 0 * 0.001 * 0.01 * 0.05 . 0.1 1Residual standard error: 19.93 on 14 degrees of freedomMultiple R-squared: 0.551, Adjusted R-squared: 0.4547F-statistic: 5.726 on 3 and 14 DF, p-value: 0.009004回归方程为 y=44.9290+1.8033x1-0.1337x2+0.1668x3（2）回归方程显著，但

6、有些回归系数不显著。（3） lm.step-step(lm.sol)Start: AIC=111.2y x1 + x2 + x3 Df Sum of Sq RSS AIC- x2 1 36.0 5599.4 109.3 5563.4 111.2- x3 1 849.8 6413.1 111.8- x1 1 4617.8 10181.2 120.1Step: AIC=109.32y x1 + x3 Df Sum of Sq RSS AIC 5599.4 109.3- x3 1 833.2 6432.6 109.8- x1 1 5169.5 10768.9 119.1 summary(lm.ste

7、p)Call:lm(formula = y x1 + x3, data = pho)Residuals: Min 1Q Median 3Q Max-29.713 -11.324 -2.953 11.286 48.679Coefficients: Estimate Std. Error t value Pr(|t|)(Intercept) 41.4794 13.8834 2.988 0.00920 *x1 1.7374 0.4669 3.721 0.00205 *x3 0.1548 0.1036 1.494 0.15592-Signif. codes: 0 * 0.001 * 0.01 * 0.

8、05 . 0.1 1Residual standard error: 19.32 on 15 degrees of freedomMultiple R-squared: 0.5481, Adjusted R-squared: 0.4878F-statistic: 9.095 on 2 and 15 DF, p-value: 0.002589x3仍不够显著。再用drop1函数做逐步回归。 drop1(lm.step)Single term deletionsModel:y x1 + x3 Df Sum of Sq RSS AIC 5599.4 109.3x1 1 5169.5 10768.9 1

9、19.1x3 1 833.2 6432.6 109.8可以考虑再去掉x3. lm.opt|t|)(Intercept) 59.2590 7.4200 7.986 5.67e-07 *x1 1.8434 0.4789 3.849 0.00142 *-Signif. codes: 0 * 0.001 * 0.01 * 0.05 . 0.1 1Residual standard error: 20.05 on 16 degrees of freedomMultiple R-squared: 0.4808, Adjusted R-squared: 0.4484F-statistic: 14.82 on

10、 1 and 16 DF, p-value: 0.001417皆显著。Ex6.3 x y plot(x,y) lm.sol summary(lm.sol)Call:lm(formula = y 1 + x)Residuals: Min 1Q Median 3Q Max-9.8413 -2.3369 -0.0214 1.0592 17.8320Coefficients: Estimate Std. Error t value Pr(|t|)(Intercept) -1.4519 1.8353 -0.791 0.436x 1.5578 0.2807 5.549 7.93e-06 *-Signif.

11、 codes: 0 * 0.001 * 0.01 * 0.05 . 0.1 1Residual standard error: 5.168 on 26 degrees of freedomMultiple R-squared: 0.5422, Adjusted R-squared: 0.5246F-statistic: 30.8 on 1 and 26 DF, p-value: 7.931e-06线性回归方程为 y=-1.4519+1.5578x，通过F 检验。 常数项参数未通过t 检验。 abline(lm.sol) y.yes y.fit y.rst plot(y.yesy.fit) pl

12、ot(y.rsty.fit)残差并非是等方差的。修正模型，对相应变量Y做开方。 lm.new summary(lm.new)Call:lm(formula = sqrt(y) x)Residuals: Min 1Q Median 3Q Max-1.54255 -0.45280 -0.01177 0.34925 2.12486Coefficients: Estimate Std. Error t value Pr(|t|)(Intercept) 0.76650 0.25592 2.995 0.00596 *x 0.29136 0.03914 7.444 6.64e-08 *-Signif. co

13、des: 0 * 0.001 * 0.01 * 0.05 . 0.1 1Residual standard error: 0.7206 on 26 degrees of freedomMultiple R-squared: 0.6806, Adjusted R-squared: 0.6684F-statistic: 55.41 on 1 and 26 DF, p-value: 6.645e-08此时所有参数和方程均通过检验。对新模型做标准化残差图，情况有所改善，不过还是存在一个离群值。第24和第28个值存在问题。Ex6.4 toothpaste lm.sol|t|)(Intercept) 4.

14、0759 0.6267 6.504 1.00e-06 *X1 1.5276 0.2354 6.489 1.04e-06 *X2 0.6138 0.1027 5.974 3.63e-06 *-Signif. codes: 0 * 0.001 * 0.01 * 0.05 . 0.1 1Residual standard error: 0.1767 on 24 degrees of freedomMultiple R-squared: 0.9378, Adjusted R-squared: 0.9327F-statistic: 181 on 2 and 24 DF, p-value: 3.33e-1

15、5回归诊断： influence.measures(lm.sol)Influence measures of lm(formula = Y X1 + X2, data = toothpaste) : dfb.1_ dfb.X1 dfb.X2 dffit cov.r cook.d hat inf1 0.00908 0.00260 -0.00847 0.0121 1.366 5.11e-05 0.16812 0.06277 0.04467 -0.06785 -0.1244 1.159 5.32e-03 0.05373 -0.02809 0.07724 0.02540 0.1858 1.283 1.

16、19e-02 0.13864 0.11688 0.05055 -0.11078 0.1404 1.377 6.83e-03 0.1843 *5 0.01167 0.01887 -0.01766 -0.1037 1.141 3.69e-03 0.03846 -0.43010 -0.42881 0.45774 0.6061 0.814 1.11e-01 0.09367 0.07840 0.01534 -0.07284 0.1082 1.481 4.07e-03 0.2364 *8 0.01577 0.00913 -0.01485 0.0208 1.237 1.50e-04 0.08239 0.01

17、127 -0.02714 -0.00364 0.1071 1.156 3.95e-03 0.046610 -0.07830 0.00171 0.08052 0.1890 1.155 1.22e-02 0.072611 0.00301 -0.09652 -0.00365 -0.2281 1.127 1.76e-02 0.073512 -0.03114 0.01848 0.03459 0.1542 1.132 8.12e-03 0.051413 -0.09236 -0.03801 0.09940 0.2201 1.071 1.62e-02 0.052214 -0.02650 0.03434 0.0

18、2606 0.1179 1.235 4.81e-03 0.095615 0.00968 -0.11445 -0.00857 -0.2545 1.150 2.19e-02 0.091016 -0.00285 -0.06185 0.00098 -0.1608 1.146 8.83e-03 0.059417 0.07201 0.09744 -0.07796 -0.1099 1.364 4.19e-03 0.173118 0.15132 0.30204 -0.17755 -0.3907 1.087 5.04e-02 0.108519 0.07489 0.47472 -0.12980 -0.7579 0

19、.731 1.66e-01 0.109220 0.05249 0.08484 -0.07940 -0.4660 0.625 6.11e-02 0.0384 *21 0.07557 0.07284 -0.07861 -0.0880 1.471 2.69e-03 0.2304 *22 -0.17959 -0.39016 0.18241 -0.5494 0.912 9.41e-02 0.102223 0.06026 0.10607 -0.06207 0.1251 1.374 5.42e-03 0.180424 -0.54830 -0.74197 0.59358 0.8371 0.914 2.13e-

20、01 0.173125 0.08541 0.01624 -0.07775 0.1314 1.249 5.97e-03 0.106926 0.32556 0.11734 -0.30200 0.4480 1.018 6.49e-02 0.103327 0.17243 0.32754 -0.17676 0.4127 1.148 5.66e-02 0.1369 source(Reg_Diag.R);Reg_Diag(lm.sol) #薛毅老师自己写的程序 residual s1 standard s2 student s3 hat_matrix s4 DFFITS s51 0.00443843 0.0

21、2753865 0.02695925 0.16811819 0.012119492 -0.09114255 -0.53021138 -0.52211469 0.05369239 -0.124367273 0.07726887 0.47112863 0.46335666 0.13857353 0.185843104 0.04805665 0.30111062 0.29532912 0.18427663 0.140368605 -0.09130271 -0.52689847 -0.51881406 0.03838430 -0.103654426 0.30162101 1.79287913 1.88

22、596579 0.09362223 0.606134067 0.03066005 0.19855842 0.19453763 0.23641540 * 0.108246268 0.01199519 0.07085860 0.06937393 0.08226537 0.020770479 0.08491891 0.49217591 0.48426323 0.04664158 0.1071124610 0.11625405 0.68315814 0.67537315 0.07261134 0.1889796911 -0.13874451 -0.81570765 -0.80983786 0.07348894 -0.2280782012 0.11540228 0.67051940 0.66263761 0.05137589 0.1542086413 0.16178406