, ( stats, ):
ctest - (, "", , , - ...)
eda - ,
lqs -
modreg :
mva -
nls
splines -
stepfun -
ts -
: library() :
> library(eda)
3.1. 2 ( ).
c2 . , . , , 5-7 ( ). , .
:
ni - , ni - ( ).
:
- , - , .
, ( ), =2 . , R chisq.test().
chisq.test (x, y = NULL, p = rep(1/length (x), length (x)))
x | . |
y | ; , x - . |
p | , x. |
x ‑ , x ‑ , y , x ‑ . , - , p, , p . x - ( ), , . x y , ( ), .
() qchisq(p,df) c2- [2,.329].
, .
|
|
N<-100 #
x.norm<-rnorm(N,mean=2,sd=2.5) #
10% ( 10 )
> x.norm.q <- quantile(x.norm,probs=seq(0,1,0.1))
> round(x.norm.q,2)
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
-4.12 -1.51 -0.14 0.59 1.52 2.15 2.70 3.89 4.51 5.22 8.15
> summary(x.norm)
Min. 1st Qu. Median Mean 3rd Qu. Max.
-1.4490 0.6675 2.0330 2.0670 3.1050 6.5830
:
> k<-6 #
> x.q <- c(-10, -1.0, 0.5, 2.0, 3.5, 5.0,12.0)
> x.norm.hist<-hist(x.norm,breaks=x.q,plot=FALSE)
> x.norm.hist$counts
12 15 22 18 18 15
( )
> x.q[1]<-(-Inf);x.q[k+1]<-(+Inf)#
> x.norm.p.theor<-pnorm(x.q,mean=mean(x.norm),sd=sd(x.norm))
> x.norm.p.theor<-(x.norm.p.theor[2:(k+1)]-x.norm.p.theor[1:k])
> round(x.norm.p.theor,2)
0.12 0.15 0.21 0.22 0.16 0.14
> chisq.test(x.norm.hist$counts,p=x.norm.p.theor)
Chi-squared test for given probabilities
data: x.norm.hist$counts
X-squared = 0.9691, df = 5, p-value = 0.965
H0 , (p -value) 0.965 (96.5%), , 96.5%, . , .
. ( ̻ ƻ), ( 0, 1).
> x<-c(̔,Ɣ,Ɣ,̔,Ɣ,Ɣ,̔,Ɣ,̔,Ɣ,Ɣ,̔,̔)
> y<-c(0, 0, 1, 0, 0, 1, 0, 0, 1, 1, 0, 1, 0)
> tbl<-table(x,y)
> tbl
y
x 0 1
4 3
4 2
() x y:
> p.x<-c(sum(tbl[1,])/sum(tbl),sum(tbl[2,])/sum(tbl))
> p.y<-c(sum(tbl[,1])/sum(tbl),sum(tbl[,2])/sum(tbl))
, ( ),
> p.x<-apply(tbl,1,sum)/sum(tbl) # 1 -
> p.y<-apply(tbl,2,sum)/sum(tbl) # 2 -
( H0), :
>p.theor<-c(p.x[1]*p.y[1],p.x[2]*p.y[1], p.x[1]*p.y[2],p.x[2]*p.y[2])
, ( ),
> p.theor<- p.x %*% t(p.y)
> chisq.test(as.vector(t),p=p.theor)
Chi-squared test for given probabilities
data: as.vector(t)
X-squared = 0.1238, df = 3, p-value = 0.9888
p -value , . , tbl , 5 ( ), - , - .