fdabl

---
title: "Bayesian effect size for ANOVA designs"
author: "Fabian Dablander, Maarten Marsman"
date: "May 22, 2016"
output: html_document
---

## 1) Example: One-way design
Download the data from [https://osf.io/zcipy/](https://osf.io/zcipy/), save it under
"example1.sav" in the current directory.

fdabl

function PCA(X)
    ## Y = PX

    # required:
    #- P such that
    #- we get rid of 2nd order correlation, Cov(Y) = D [some diagonal matrix]

    # method:
    #- project onto an orthogonal space spanned by the
    #- eigenvectors with variances equal to the eigenvalues

fdabl

library('rvest')
library('stringr')


fachschaft = 'http://wiki.psyfako.org/index.php?title=Liste_der_Psychologie_Fachschaften'

emails = read_html(fachschaft) %>%
  html_nodes('p, a, span') %>% html_text(.) %>% 
  str_replace('\\[at\\]', '@') %>% str_replace('\\(at\\)', '@') %>%
  str_replace('\\[ät\\]', '@') %>% str_replace('\\[dot\\]', '.') %>% 

fdabl

shinyApp(
  options = list(width = '25%', height = '25%'),
  ui = shinyUI(fluidPage(
    sidebarLayout(
      sidebarPanel(
        sliderInput('e', label = 'e', min = 0, max = 100, value = 10, step = 1),
        sliderInput('N', label = 'N', min = 10, max = 500, value = 100, step = 1),
        numericInput('seed', label = 'random seed', value = 1774),
        numericInput('lambda', label = 'penalty (lambda)', value = 0),
        textInput('true', label = 'true function f(x)', value = 'x^2'),

fdabl

library('runjags')


lasso.bayes <- function(y, X, n.iter = 10000) {
    ms <- '
    model {
        sigma ~ dunif(0, 100)
        tau <- pow(sigma, -2)

        lambda ~ dunif(0, 10)

fdabl

# c('practicality', 'difficulty', 'some color, dunno')

papers <- rbind(
# paper we discuss
c(5, 3, 1), # Lindley (1993)
c(8, 0, 1), # Kruschke (2010)
c(4, 3, 1), # Wagenmakers (2007)
c(5, 2, 1), # Dienes (2011)
c(3, 7, 1), # Vandekerckhove, Matzke, Wagenmakers (2015)
c(4, 8, 1), # Rouder & Morey (2012)

fdabl

library('shiny')
# the code (with a little cleaning up) for the visualisations is from
# http://alexanderetz.com/2015/07/25/understanding-bayes-updating-priors-via-the-likelihood/
# Alex Etz runs a really nice blog -- go check it out!

shinyApp(
  ui = shinyUI(fluidPage(
    sidebarLayout(
      sidebarPanel(
        div(style = 'display: inline-block;', numericInput('a', label = h4('a'), value = 1)),

fdabl

#' False Discovery Rate (FDR)
#' see also http://www.sciencedirect.com/science/article/pii/S1053811901910377
#'
#'          Declared active  Declared inactive
#' Active       V_aa               V_ai       T_a
#' Inactive     V_ia               V_ii       T_i
#'              D_a                D_i         V
#'
#' instead of controlling the overall family wise error, we control
#' the proportion of false discoveries; that is, we want (V_aa) / (V_aa + V_ia)

fdabl

library('shiny')
# read this: http://alexanderetz.com/2015/04/15/understanding-bayes-a-look-at-the-likelihood/


shinyApp(
  ui = shinyUI(fluidPage(
    sidebarLayout(
      sidebarPanel(
        sliderInput("p1", label = "p1",
                    min = 0, max = 1, value = 0.7, step = 0.01),

fdabl

/*
 * empirically tests the complexity of a bogosort based algorithm which
 * shuffles circles until they don't overlap (for stimulus presentation)
 * is a function of the circle radius, the canvas width and height
 *
 * call it with (first arg is radius, then width and height of canvas)
 * > node test.js 10 300 500
 * > node test.js 10
 * etc.
 */