# provides skew(), kurtosi()
require(psych)

multiplot <- function(row,col) {
# Lays out the current graphics device in a grid
#   no-frills, described in Vasishth & Broe (2010)
#
# ARGS: row, number of rows in the grid
#       col, number of cols in the grid
#
# RETURN: none
  
  # square plotting regions, layed out row-by-row
  par(mfrow = c(row,col), pty="s")

  # sets the plotting margin, bottom-left-top-right
  par(mar=c(3,3,3,2))

}

simulateExperiment<-function(my.population, n.obs = 40, n.sim = 100000){
# Simulate any number of 'experiments', of any sample size
# With any vector data
#
# ARGS: n.obs - number of observations per experiment/sample
#       n.sim - number of experiments/samples
#
# RETURN: vector of  means for the series of sample
  
  mean.sample <- vector("numeric", length=n.sim)
  
  for(i in 1:n.sim){
    sample(my.population, size=n.obs, replace=TRUE) -> rt.sample
    mean(rt.sample) -> mean.sample[i]
  }
  
  return(mean.sample)
  
}

se<-function(x) {
	y <- x[!is.na(x)]
	sqrt(var(as.vector(y))/length(y))
}

###
### 95% CONFIDENCE INTERVAL
###

ci<-function(x, size=0.95) {
#
# ARGS:
# 
# RETURNS: 

	m<-mean(x,na.rm=TRUE)
	standard.error<-se(x)
	len<-length(x)
	upper<-m+qt(size+(1-size)/2,df=len-1)*standard.error
	lower<-m+qt((1-size)/2,df=len-1)*standard.error
	return(data.frame(lower=lower,upper=upper))

}