even4void

(require 'offlineimap)
(add-to-list 'load-path "~/.emacs.d/lib/mu4e")
(require 'mu4e)
(require 'mu4e-maildirs-extension)
(mu4e-maildirs-extension)

(setq mu4e-drafts-folder "/drafts"
      mu4e-sent-folder   "/sent"
      mu4e-trash-folder  "/trash")
(setq mu4e-maildir-shortcuts

even4void

## ESTIMATE SAMPLE SIZE NEEDED to achieve a given power in univariate twin studies.
##
## Input:
##   pmz: sampling fraction MZ/DZ
##   tmz: intraclass correlation for MZ twins
##   tdz: intraclass correlation for DZ twins
##   alpha: level I risk level
##   beta: level II risk level (power = 1 - beta)
##
## Output:

even4void

library(lme4)
data(sleepstudy)

## Fit individual regression lines for each subject
dfrm <- coef(lmList(Reaction ~ Days | Subject, sleepstudy))

## Estimate parameters of a random intercept and random intercept and slope model
m1 <- lmer(Reaction ~ Days + (1 | Subject), data=sleepstudy)
m2 <- lmer(Reaction ~ Days + (Days | Subject), data=sleepstudy)

even4void

RMDFILE=demo-rmd-pandoc
PANDOC=~/.cabal/bin/pandoc

all:
	Rscript -e "require(knitr); require(markdown); knit('$(RMDFILE).rmd', '$(RMDFILE).md'); purl('$(RMDFILE).rmd')"
	${PANDOC} --mathjax --toc -B header.html -A footer.html --bibliography refs.bib --css markdown.css -s $(RMDFILE).md -o $(RMDFILE).html
 

even4void

# Compute % agreement between any two columns in a data.frame
dfrm <- data.frame(replicate(18, sample(letters[1:3], 30, replace=TRUE)))

agree <- function(df, i, j) {
  tab <- table(df[,c(i,j)])   
  sum(diag(tab))/sum(tab)
}

# Solution 1
vagree <- Vectorize(agree, vectorize.args=list("i","j"))

even4void

#! /usr/bin/env sh

# Post-process Stata do -> log file.

# We should ensure that we use the correct stata program
# (e.g., might be stata, stata-mp, etc.). I'll put this here 
# for later update.
STATA="$(which stata)"

# Process command line options

even4void

# simulate some data
set.seed(88)
n <- 30
x <- runif(n)
y <- 1.2 + 0.8*x + rnorm(n)
g <- sample(letters[1:2], n, replace=TRUE)
my.df <- data.frame(x=x, y=y, grp=factor(g))

# show data structure
dput(my.df)

even4void

mha <- 1  # mean under the alternative
es <- 2   # observed effect (deviation from mean under H0)
x <- seq(-6, 6, length=1000)
dh0 <- dnorm(x, 0, 1)

show.it <- function(es, mha, verbose=FALSE) {
  dh1 <- dnorm(x, mha, 1)

  plot.new()
  plot.window(xlim=range(x), ylim=c(0,.6)) 

even4void

tufte.boxplot <- function(x, g) {
  k <- nlevels(g)
  crit.val <- tapply(x, g, median)
  plot(1:k, crit.val, ylim=c(min(x)*1.1, max(x)*1.1), pch=19,
       xlab=deparse(substitute(g)), ylab=deparse(substitute(x)))
  for (i in 1:k) {
    tmp <- boxplot.stats(x[as.numeric(g)==i])
    segments(i, tmp$stats[1], i, tmp$stats[2])
    segments(i, tmp$stats[4], i, tmp$stats[5])
    points(rep(i, length(tmp$out)), tmp$out, cex=.8)

even4void

@article{Narrow2009,
	Author = {Narrow, W.E. and Kuhl, E.A. and Regier, D.A.},
	Journal = {World Psychiatry},
	Month = {Jun},
	Number = {2},
	Pages = {88-89},
	Title = {{DSM-V} perspectives on disentangling disability from
                  clinical significance},
	Volume = {8},
	Year = {2009}}