r:drawing_sampling_distribution_plot
This is an old revision of the document!
rm(list=ls())
rnorm2 <- function(n,mean,sd){
mean+sd*scale(rnorm(n))
}
n.p <- 10000
m.p <- 100
sd.p <- 10
p1 <- rnorm2(n.p, m.p, sd.p)
m.p1 <- mean(p1)
sd.p1 <- sd(p1)
p2 <- rnorm2(n.p, m.p+10, sd.p)
m.p2 <- mean(p2)
sd.p2 <- sd(p2)
n.s <- 100
se.z <- c(sqrt(var(p1)/n.s))
x_values <- seq(mean(p1)-5*se.z,
mean(p1)+15*se.z,
length.out = 500)
# Calculate the probability
# density for a normal distribution
y_values <- dnorm(x_values,
mean = mean(p1),
sd = se.z)
# Plot the theoretical PDF
plot(x_values, y_values, type = "l",
lwd=3,
main = "Distribution of Sample Means",
xlab = "Value", ylab = "Density")
m.p1 <- mean(p1)
se1 <- c(m.p1-se.z, m.p1+se.z)
se2 <- c(m.p1-2*se.z, m.p1+2*se.z)
se3 <- c(m.p1-3*se.z, m.p1+3*se.z)
abline(v=c(m.p1,se1,se2,se3),
col=c('black', 'red', 'red',
'green', 'green',
'blue', 'blue'),
lwd=2)
treated.s <- sample(p2, n.s)
m.treated.s <- mean(treated.s)
abline(v=m.treated.s, col='orange', lwd=2)
diff <- m.treated.s-mean(p1)
diff/se.z
# usual way - using sample's variance
# instead of p1's variance to get
# standard error value
se.s <- sqrt(var(treated.s)/n.s)
se.s
diff/se.s
pt(diff/se.s, df=n.s-1, lower.tail = F) * 2
t.test(treated.s, mu=m.p1, var.equal = T)
r/drawing_sampling_distribution_plot.1757542962.txt.gz · Last modified: by hkimscil