除非源码出现问题，请勿回复
对于不屈的分析，帖子在这里：
https://tieba.baidu.com/p/6418940497
This code is under MIT lisence. Please cite author's name (Hedonist55) when reference the code.
不屈实验源码 （peach --> 桃的数量，trial ---> 实验次数，返回一个数组每个数表示在某次实验周泰在第几张屈死掉的）
indomitability.experiment <- function(peach, trial) {
deck <- rep(1:13, each=8)
result <- 1:trial
for (i in 1:trial) {
experiment.count <- rep(FALSE, 13)
deck.shuffle <- sample(deck)
end <- FALSE
index <- 1
peach.used <- 0
while (!end) {
point <- deck.shuffle[index]
if(!experiment.count[point]) {
experiment.count[point] <- TRUE
index <- index + 1
} else {
if(peach.used == peach) {
end <- TRUE
} else {
peach.used <- peach.used + 1
index <- index + 1
}
}
}
result[i] <- index
}
return(result)
}

理论模型计算源码：
叫emprical就好，peach指的是桃的数量（0到7之间）
conditional.prob <- function(x, y) {
result <- 0
for (i in 1:(x - 1 - y)) {
part.log <- log(choose(x - 1 - y, i)) + log(choose(8 * i, x - 1)) - log(choose(104, x - 1))
part <- exp(part.log) * ((-1) ^ (x - 1 - y - i))
result <- result + part
}
return(result)
}
emprical <- function(peach) {
result <- 1:13
sample.space <- (peach + 2):(peach + 14)
for (i in 1:13) {
card <- sample.space[i]
cur <- conditional.prob(card, peach) * choose(13, card - peach - 1)
cur <- cur * (8 * (card - peach - 1) - (card - 1)) / (104 - (card - 1))
result[i] <- cur
}
names(result) <- sample.space
return(result)
}

这个是R码，需要下载R才能正确食用。

R语言？