Prvi zadatak
\(X\sim \mathcal{U}(0,1)\). Odrediti raspodelu:
- \(Y = aX+B\)
- \(W = -\log X\)
1. \(Y = aX+B\)
Od ranije, treba da bude \(Y\sim\mathcal{U}(b, a+b)\).
x <- runif(1e4)
hist(x)
y <- 3 * x + 7
hist(y)
Raspodela \(W=-\log X\)
Treba da bude \(\mathcal{E}(1)\).
hist(x)
w <- -log(x)
hist(w)
# curve uvek prima izraz po x
curve(dexp(x), xlim = c(0, 5))
hist(w, probability = TRUE)
curve(dexp(x), xlim = c(0, 5), add = TRUE, col = "red")
Drugi zadatak
\(X \sim \mathcal{E}(\lambda)\). Odrediti raspodele:
- \(Y = [X]\)
- \(W = 1-e^{-\lambda X}\)
Raspodela \(Y\)
Treba da bude \(\mathcal{G}(1-e^{-\lambda})\)
x <- rexp(1e4, rate = 0.1) #lambda = pi
hist(x, prob = TRUE)
y <- floor(x)
hist(y, prob = TRUE)
pts <- seq_len(50)
points(dgeom(pts, prob = 1-exp(-0.1)), col = "red")
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