16장 비율(Proportions)

작성자안재형|작성시간09.12.15|조회수601 목록 댓글 0

# H0: p=1/2 vs Ha: p is not 1/2

# 44 heads out of 100 trials

 

### Exact binomal test

> binom.test(44,100)

 

        Exact binomial test

data:  44 and 100
number of successes = 44, number of trials = 100, p-value = 0.2713
alternative hypothesis: true probability of success is not equal to 0.5
95 percent confidence interval:
 0.3408360 0.5428125
sample estimates:
probability of success
                  0.44

 

### Normal approximation

> prop.test(44,100)

 

        1-sample proportions test with continuity correction

data:  44 out of 100, null probability 0.5
X-squared = 1.21, df = 1, p-value = 0.2713
alternative hypothesis: true p is not equal to 0.5
95 percent confidence interval:
 0.3420219 0.5426412
sample estimates:
   p
0.44

 

# 연속성 수정을 안하려면

> prop.test(44,100,correct=FALSE)

        1-sample proportions test without continuity correction

data:  44 out of 100, null probability 0.5
X-squared = 1.44, df = 1, p-value = 0.2301
alternative hypothesis: true p is not equal to 0.5
95 percent confidence interval:
 0.3467203 0.5377190
sample estimates:
   p
0.44

 

##### H0: p1 = p2 vs Ha: p1 is not p2

> prop.test(c(9,8),c(50,60))

 

        2-sample test for equality of proportions with continuity correction

data:  c(9, 8) out of c(50, 60)
X-squared = 0.1676, df = 1, p-value = 0.6823
alternative hypothesis: two.sided
95 percent confidence interval:
 -0.1085548  0.2018881
sample estimates:
   prop 1    prop 2
0.1800000 0.1333333

 

##### 적합성 검정(goodness of fit)

# Mendel's law; 9:3:3:1

> chisq.test(c(216,79,65,21), p=c(9,3,3,1)/16)

        Chi-squared test for given probabilities

data:  c(216, 79, 65, 21)
X-squared = 1.7262, df = 3, p-value = 0.6311

 

##### 두 비율의 차이

 

> prop.test(c(9,8),c(50,60))

 

        2-sample test for equality of proportions with continuity correction

data:  c(9, 8) out of c(50, 60)
X-squared = 0.1676, df = 1, p-value = 0.6823
alternative hypothesis: two.sided
95 percent confidence interval:
 -0.1085548  0.2018881
sample estimates:
   prop 1    prop 2
0.1800000 0.1333333

 

##### 셋 이상 비율의 차이

> prop.test(c(14,18,15),c(100,100,100))

        3-sample test for equality of proportions without continuity
        correction

data:  c(14, 18, 15) out of c(100, 100, 100)
X-squared = 0.656, df = 2, p-value = 0.7204
alternative hypothesis: two.sided
sample estimates:
prop 1 prop 2 prop 3
  0.14   0.18   0.15

 

##### Simulation

# 이항분포

> set.seed(1234)
> rbinom(n=10,size=1,p=0.5)
 [1] 0 1 1 1 1 1 0 0 1 1
> set.seed(1234)
> rbinom(n=1,size=10,p=0.5)
[1] 3
> set.seed(1234)
> rbinom(n=5,size=10,p=0.5)
[1] 3 5 5 6 7

 

# 다항분포

> set.seed(1234)
> rmultinom(n=1, size=1, p=c(1,1,1,1,1,1)/6)
     [,1]
[1,]    0
[2,]    0
[3,]    0
[4,]    0
[5,]    1
[6,]    0
> set.seed(1234)
> rmultinom(n=1, size=10, p=c(1,1,1,1,1,1)/6)
     [,1]
[1,]    0
[2,]    2
[3,]    2
[4,]    2
[5,]    3
[6,]    1
> set.seed(1234)
> rmultinom(n=5, size=10, p=c(1,1,1,1,1,1)/6)
     [,1] [,2] [,3] [,4] [,5]
[1,]    0    2    2    3    1
[2,]    2    0    2    1    1
[3,]    2    1    1    1    1
[4,]    2    3    3    1    0
[5,]    3    2    1    1    2
[6,]    1    2    1    3    5

 

# 멘델의 강낭콩 교배 실험

 

> set.seed(1234)
> rmultinom(n=1, size=381, p=c(9,3,3,1)/16)
     [,1]
[1,]  222
[2,]   66
[3,]   68
[4,]   25

 

 

다음검색
현재 게시글 추가 기능 열기

댓글

댓글 리스트
맨위로

카페 검색

카페 검색어 입력폼