多元統計分析與R語言建模複習筆記
參考資料：
多元統計分析及R語言建模
王斌會教授
視頻：
https://www.icourse163.org/learn/JNU-1002335007#/learn/content?type=detail&id=1007583075&sm=1
資源：
http://rstat.leanote.com/cate/多元統計分析
4. 多元相關與迴歸分析及R使用
> x=c(171,175,159,155,152,158,154,164,168,166,159,164)
> y=c(57,64,41,38,35,44,41,51,57,49,47,46)
> plot(x,y)
> cor(x,y)
[1] 0.9593031
> cor.test(x,y) # 相關係數的假設檢驗
Pearson's product-moment correlation
data: x and y
t = 10.743, df = 10, p-value = 8.21e-07
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
0.8574875 0.9888163
sample estimates:
cor
0.9593031
迴歸係數的假設檢驗
> d4.3=read.table('clipboard',header = T)
> d4.3
y x
1978 11.3262 5.1928
1979 11.4638 5.3782
1980 11.5993 5.7170
1981 11.7579 6.2989
1982 12.1233 7.0002
1983 18.6695 7.5559
1984 16.4286 9.4735
1985 20.0482 20.4079
1986 21.2201 20.9073
1987 21.9935 21.4036
1988 23.5724 23.9047
1989 26.6490 27.2740
1990 29.3710 28.2187
1991 31.4948 29.9017
1992 34.8337 32.9691
1993 43.4895 42.5530
1994 52.1810 51.2688
1995 62.4220 60.3804
1996 74.0799 69.0982
1997 86.5114 82.3404
1998 98.7595 92.6280
1999 114.4408 106.8258
2000 133.9523 125.8151
2001 163.8604 153.0138
2002 189.0364 176.3645
2003 217.1525 200.1731
2004 263.9647 241.6568
2005 316.4929 287.7854
2006 387.6020 348.0435
2007 513.2178 456.2197
2008 613.3035 542.1962
> m4.3=lm(y~x,data=d4.3)
> m4.3
Call:
lm(formula = y ~ x, data = d4.3)
Coefficients:
(Intercept) x
-1.197 1.116
> plot(y~x,data=d4.3)
> abline(m4.3) # 添加迴歸線
> summary(m4.3) # 迴歸方程的假設檢驗
Call:
lm(formula = y ~ x, data = d4.3)
Residuals:
Min 1Q Median 3Q Max
-6.630 -3.692 -1.535 5.338 11.432
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -1.19656 1.16125 -1.03 0.311
x 1.11623 0.00674 165.61 <2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Residual standard error: 5.095 on 29 degrees of freedom
Multiple R-squared: 0.9989, Adjusted R-squared: 0.9989
F-statistic: 2.743e+04 on 1 and 29 DF, p-value: < 2.2e-16
6.1 線性判別分析
> d6.1 = read.table('clipboard',header = T)
> d6.1
G x1 x2
1 1 -1.9 3.2
2 1 -6.9 0.4
3 1 5.2 2.0
4 1 5.0 2.5
5 1 7.3 0.0
6 1 6.8 12.7
7 1 0.9 -5.4
8 1 -12.5 -2.5
9 1 1.5 1.3
10 1 3.8 6.8
11 2 0.2 6.2
12 2 -0.1 7.5
13 2 0.4 14.6
14 2 2.7 8.3
15 2 2.1 0.8
16 2 -4.6 4.3
17 2 -1.7 10.9
18 2 -2.6 13.1
19 2 2.6 12.8
20 2 -2.8 10.0
> attach(d6.1)
The following objects are masked _by_ .GlobalEnv:
x1, x2
> plot(x1,x2)
> plot(d6.1$x1,d6.1$x2)
> library(MASS)
> ld = lda(G~x1+x2)
Error in model.frame.default(formula = G ~ x1 + x2) :
變數的長度不一樣('x1')
> ld = lda(G~d6.1$x1+d6.1$x2)
> ld
Call:
lda(G ~ d6.1$x1 + d6.1$x2)
Prior probabilities of groups:
1 2
0.5 0.5
Group means:
d6.1$x1 d6.1$x2
1 0.92 2.10
2 -0.38 8.85
Coefficients of linear discriminants:
LD1
d6.1$x1 -0.1035305
d6.1$x2 0.2247957
> lp=predict(ld)
> lp
$class
[1] 1 1 1 1 1 2 1 1 1 1 2 2 2 2 1 2 2 2 2 2
Error in if (n <= 1L || lenl[n] <= width) n else max(1L, which.max(lenl > :
missing value where TRUE/FALSE needed
> lp$class
[1] 1 1 1 1 1 2 1 1 1 1 2 2 2 2 1 2 2 2 2 2
Levels: 1 2
> data.frame(G,lp$class)
G lp.class
1 1 1
2 1 1
3 1 1
4 1 1
5 1 1
6 1 2
7 1 1
8 1 1
9 1 1
10 1 1
11 2 2
12 2 2
13 2 2
14 2 2
15 2 1
16 2 2
17 2 2
18 2 2
19 2 2
20 2 2
快速聚類法
> d7.2=read.table('clipboard',header = T)
> plot(d7.2)
> install.packages("D:/Programing/多元統計分析與R語言/例題數據/mvstats.zip", repos = NULL, type = "win.binary")
Installing package into ‘C:/Users/Lenovo/Documents/R/win-library/3.5’
(as ‘lib’ is unspecified)
package ‘mvstats’ successfully unpacked and MD5 sums checked
> library(mvstats)
> H.clust(d7.2,m='single',plot=T)
Call:
hclust(d = D, method = m)
Cluster method : single
Distance : euclidean
Number of objects: 31
