guidocor

# Helper function 
# gets a fitted glmer model 
# Example:
# m0 <- glmer(acc ~ cond + (0 + rt | subject_nr) + (1 | item) + (1 | subject_nr), family=binomial(link=logit), data = datos.x.rt)
# Wald is fast and innacurate, better try the longer "profile"


plot.glmer <- function(m0, my.method = "Wald") {
  # that could take a while 
  confint_m0 <- confint(m0, level = 0.95, method=my.method)

guidocor

#---------------------------------------------------------------
# INSTALL MY PACKAGES SNIPPET install.R
#---------------------------------------------------------------

# Script to detach all previous packages loaded, install those who you
# use and import them. Change the needed argument to install others or paste to your script
# the install_my_packages() fucntion and give other argument

#
# USEFUL:

guidocor

guidocor

# -------------------------------------------------------
# READ IMAGES: Script that reads the images (png) of a folder and
# create a csv with number, size y bytes, x and y length
# -------------------------------------------------------
setwd("~/Dropbox/p1/")
source("./utils/install.R") #source("https://raw.githubusercontent.com/guidocor/R_utils/master/install.R")

install_and_detach("png",clean = F,load = T)
# my own script (install.R), but you can use
# install.packages("png"); library(png)

guidocor

# Cleaning the session
rm(list=ls());gc()
# Useul function to load and install al needed packages
# we are reading it from my gist on GitHub.
source("https://raw.githubusercontent.com/guidocor/R_utils/master/install.R")
install_and_detach(c("dplyr","ggplot2", "doBy"),load= T, clean = T)

theme_set(theme_bw()) # My favourite ggplot2 theme : )  
parts = 30 # Our participants
trials = 30 # Our Trials

guidocor

# helper function that makes a correlation 
# Computes correlation between x and y and returns a list with the
# correlation object and the text as APA states to report it
# Also, the function prints the result, you can change this behavior with silent = FALSE
# Correlation reporting in APA Style: http://my.ilstu.edu/~jhkahn/apastats.html
# Brackets of CI in APA Style: http://blog.apastyle.org/apastyle/2010/06/formatting-statistics-using-brackets.html
correlation <- function(x,y, silent = TRUE, my.method="pearson") {

guidocor

iqr_1.5 <- function(df, rt, conds, FUN = function(x) { 
  c(iqr = IQR(x),
    lower = quantile(x, c(0.25))-1.5*(quantile(x, c(0.75))-quantile(x, c(0.25))),
    upper = quantile(x, c(0.75))+1.5*(quantile(x, c(0.75))-quantile(x, c(0.25)))) 
}){
  # This function calculates the 1.5 * IQR of each participant and condition
  # returns a dataframe with the columns formatted.
  # Is used to detect outliers in reaction times at participant per condition level.
  # Also, yo can use it with custom functions

guidocor

###"Plotting model estimates in base R graphics"


#There are many ways of displaying data and graphs in R, 
#here I propose a base graph solution to plot estimates and confidence 
#intervals using a Generalized Linear Mixed Effect Model, but this code 
#could be used for other kind of models. So, we have 
#a Generalized Linear Effect Model and we want to plot the 
#model estimates and its confidence interval. First we install 

guidocor

# Snippet to call and install all libraries required with pacman package
if (!require('pacman')) install.packages('pacman'); library('pacman') 
p_load(tidyverse, yarrr, trimr, lattice, doBy, Rmisc, reshape2)

# Dataset taken from trimr package. Check it out and 
# don't miss this other way to do Reaction Time outlier detection
# at https://cran.r-project.org/web/packages/trimr/vignettes/trimr-vignette.html
data(exampleData)
head(exampleData)  

guidocor

#############################################################
## Estimating and plotting Generalized Linear Mixed Models ##
#############################################################

# This script will
# 1) Simulate some experimental data
# 2) Run a glmer with afex
# 3) Estimate predicted marginal means with emmeans
# 4) Plot estimates