fditraglia

#This script carries out a simulation study similar to that in Section 2 of ``Posterior Predictive Evidence on US Inflation...'' by Basturk et al (2012), exploring the effects of trend mis-specification on the estimate of the persistence of the inflation parameter. 

#Library to implement Bai and Perron (2003)
library(strucchange)

set.seed(4513)

g <- 0.5 #Inflation persistence
N.sims <- 250 #Simulation repetitions

fditraglia

library(Quandl)
p <- 100 * Quandl('FRED/GDPDEF', collapse = 'quarterly', start_date = '1960-01-01', end_date = '2012-01-01', type = 'ts', transformation = 'rdiff')

#Michigan Inflation Expectations aren't in Quandl yet
library(quantmod)
getSymbols("MICH",src="FRED")
#Median expected price change next 12 months, Survey of Consumers.

#Convert to ts object
MICH.start <- as.numeric(format.Date(start(MICH), '%Y'))

fditraglia

x <- 5
x == 5



foo <- 3
bar <- 5
foo.bar <- foo + bar
foo.bar2 <- 2 * foo.bar

fditraglia

#1.1 Expressions 
1 + 1
"Arr, matey!"
6*7

#1.2 Logical Values
3 < 4
2 + 2 == 5

#1.3 Variables

fditraglia

survey <- read.csv("http://www.ditraglia.com/econ103/old_survey.csv")



survey <- survey[,1:6]

fditraglia

survey <- read.csv("http://www.ditraglia.com/econ103/old_survey.csv")



survey <- survey[,c("handedness", "height", "handspan")]
head(survey)

fditraglia

marbles <- c('red', 'blue', 'green')
sample(x = marbles, size = 2, replace = FALSE)



sample(x = 1:10, size = 5, replace = FALSE)

fditraglia

#This function simulates n independent, identically distributed Bernoulli Random Variables
rbern <- function(n, p){
  
  sims <- sample(0:1, size = n, replace = TRUE, prob = c(1-p, p))
  return(sims)
  
}

fditraglia

x <- seq(from = -1, to = 1, by = 0.5)
y <- x^2
plot(x, y)



x <- seq(from = -1, to = 1, by = 0.1)
y <- x^2
plot(x, y)

fditraglia

x <- seq(from = -5, to = 5, by = 0.01)
y1 <- dnorm(x)
y2 <- dt(x, df = 1)
y <- cbind(y1, y2)
matplot(x, y, lty = 1, type = 'l', ylab = 'f(x)')