[統算] chisq.test
# chisq.test.r
- rescale.p = FALSE, simulate.p.value = FALSE, B = 2000)
+ rescale.p = FALSE, simulate.p.value = FALSE, B = 2000, unspec.m = 0)
else {
- PARAMETER <- length(x) - 1
+ if(unspec.m){
+ PARAMETER <- length(x) - unspec.m - 1
+ }
+ else {
+ PARAMETER <- length(x) - 1
+ }
PVAL <- pchisq(STATISTIC, PARAMETER, lower.tail = FALSE)
}
## Example 9c ##
> N <- c(6, 2, 1, 9, 7, 5)
> prob <- c(0.0550, 0.1596, 0.2312, 0.2237, 0.1622, 0.1682)
> chisq.test(N, p=prob, unspec.m=1)
Chi-squared test for given probabilities
data: N
X-squared = 19.887, df = 4, p-value = 0.0005257
Warning message:
In chisq.test(N, p = p, unspec.m = 1) :
Chi-squared approximation may be incorrect
--
改一下 chisq.test 的函式，加入一個 unspec.m（unspecified parameter 的個數）
這樣跑出來的結果就跟 Example 9c 一樣了，也就是我們要的 goodness of fit :p
--
※ 發信站: 批踢踢實業坊(ptt.cc)
◆ From: 140.115.221.212
※ 編輯: Chris7462 來自: 140.115.221.212 (06/04 01:25)
推
06/04 12:47, , 1F
06/04 12:47, 1F
推
06/05 01:25, , 2F
06/05 01:25, 2F
推
06/05 14:58, , 3F
06/05 14:58, 3F