The CDF of y = the minimum of n Uniform(a,b) continuous random variables is:
\[ \begin{equation} F(y) = 1 - \Big(\dfrac{b-y}{b-a}\Big)^{n} \end{equation} \] We are interested in the discrete case of y = the minimum of n Uniform(1, k) discrete random variables. The CDF is:
\[ \begin{equation} F(y) = 1 - \Big(1 - \dfrac{y}{k}\Big)^{n} \end{equation} \]
By recognizing that the above formula is an accumulation function, we can use it to get the Probability Mass Function of the discrete case:
\[ \begin{equation} f(y) = \Big(1 - \dfrac{y-1}{k}\Big)^{n} - \Big(1 - \dfrac{y}{k}\Big)^{n} \end{equation} \]
“Your organization owns a machine that has a manufacturer’s expected lifetime of 10 years. This means that we expect one failure every 10 years. What is the probability that the machine will fail after 8 years?”
\[ \begin{equation} P(x>8) = \Big(1 - \dfrac{1}{10}\Big)^{8} \end{equation} \]
p <- 1/10
1-sum(dgeom(0:7, p)) #the R function starts at 0, so must do 0:7 instead of 1:8
## [1] 0.4304672
1/p #mean
## [1] 10
sqrt((1-p)/p^2) #sd
## [1] 9.486833
\[ \begin{equation} P(x>=8) = e^{-8/10} \end{equation} \]
lambda <- 1/10
pexp(8, lambda, lower.tail = F)
## [1] 0.449329
1/lambda #mean
## [1] 10
1/lambda #sd
## [1] 10
\[ \begin{equation} P(x=0) = \Big(\dfrac{9}{10}\Big)^{8} \end{equation} \]
p <- 1/10
n <- 8
dbinom(0, n, p)
## [1] 0.4304672
n*p #mean
## [1] 0.8
sqrt(p*(1-p)*n) #sd
## [1] 0.8485281
\[ \begin{equation} PMF = \dfrac{\lambda^{k}e^{-\lambda}}{k!} \end{equation} \]
p <- 1/10
n <- 8
lambda <- n*p
dpois(0, lambda)
## [1] 0.449329
lambda #mean
## [1] 0.8
sqrt(lambda) #sd
## [1] 0.8944272