install.packages(MASS)
library(MASS)
# a) B + C < 1/2
knitr::opts_chunk$set(echo = TRUE)
B <- runif(1000, min = 0, max = 1)
C <- runif(1000, min = 0, max = 1)
X <- B + C
plot(density(X))
shapiro.test(X)
##
## Shapiro-Wilk normality test
##
## data: X
## W = 0.99114, p-value = 1.02e-05
qqnorm(X)
qqline(X, col = 2)
hist(X)
# I utalized the shapiro-wilk test and qqplots to show that X is a normal distribution.
# so according to the graph of X distribution, the X < 1/2 is equal to 0.5/2 = 0.125
# b) BC < 1/2
Y <- B * C
fit1 <- fitdistr(Y, "exponential")
hist_Y <- hist(Y, freq = FALSE, breaks = 10, xlim = c(0, quantile(Y, 0.99)))
curve(dexp(x, rate = fit1$estimate), col = "red", add = TRUE)
hist_density <- sum(hist_Y$density * .1)
hist_density_0.5 <- sum(hist_Y$density[1:5]*.1)
hist_density_0.5
## [1] 0.838
# The Y distribution is exponential distribution.
#c) |B -C| < 1/2
Z <- abs(B-C)
hist(Z)
#From the plot, we can estimate the function f(x) = -200x + 200
int.fun <- function(x){-200*x + 200}
Pro_0.5 <- integrate(int.fun, lower = 0, upper = 0.5)
Pro_0.5
## 75 with absolute error < 8.3e-13
\[\\P(max\{B,C\}) = P(B\leqslant\frac{1}{2}) * P(C\leqslant\frac{1}{2}) \\= 0.5 * 0.5 \\= 0.25 \]
\[\\P(min\{B,C\}) = 1 - P(max\{B,C\}) \\= 1 - 0.25 \\= 0.75 \]