jvcasillas

\tikzstyle{every picture}+=[remember picture]
\everymath{\displaystyle}

\begin{frame}[m]\frametitle{}

\tikzstyle{na} = [baseline=-.5ex]

\begin{itemize}[<+-| alert@+>] % makes everything red
    \item Sujeto
        \tikz[na] \node[coordinate] (n1) {};

jvcasillas

\begin{frame}[m]\frametitle{}
\begin{tikzpicture}
\node[] (image) at (0,0) {\includegraphics[scale=.27]{btard_ae}\includegraphics[scale=.27]{btemp_ae}\includegraphics[scale=.27]{mono_ae}};
\end{tikzpicture}
\end{frame}

\begin{frame}[m]\frametitle{}
\begin{tikzpicture}
\node[] (image) at (0,0) {\includegraphics[scale=.27]{btard_ae}\includegraphics[scale=.27]{btemp_ae}\includegraphics[scale=.27]{mono_ae}};
\draw [-latex] (-5.35,.5) to[bend right=15] (-4.45,-0.15);

jvcasillas

# A couple things about this script...
# All text that comes after a "#" sign implies a comment that will not
# be evaluated by R. All you need to do is copy and paste the contents 
# of this script into R and it will perform an ANOVA and make a graph.
# It is important that your data (in your data file from the hw) are 
# in the correct format. You need to have the following headers for 
# each column: "participant", "ending", "stimulus", "proportion".
# Check the example file "tarea5.txt" if you need to compare.

# The following command reads the data file into R.

jvcasillas

library(ggplot2) 
plot <- ggplot(heat_plot, aes(fpro, dpro))
plot + geom_tile(aes(fill=mean_choice), colour="white") + 
     scale_fill_gradient(low="white", high="steelblue")

jvcasillas

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Generalized Linear Models
with binary categorical data
with binomial error structure
and logit link function

Categorical data is heteroskedastic. Residuals are not normally distributed. It violates assumptions of linear models. It can be approximated as a logistic curve (s-like-curve). This curve represents probability values: probability of observing 1 or 0 at y when steps in x change. (Ideal for perception data.)

model = glm(response ~ step in continuum or Hz, data=study, family="binomial") # regular generalized (logistic) linear model

jvcasillas

# Plot Spectrum Proportion 
my_data$group <- factor(my_data$group, levels = c("EL", "NE", "LL"))
df<-with(my_data, aggregate(fpro, list(group=group, fstim=fstim), mean))
df$se<-with(my_data, aggregate(fpro, list(group=group, fstim=fstim), function(x) sd(x)/sqrt(10)))[,3]
gp<-ggplot(df, aes(x=fstim, y=x, colour=group, ymin=x-se, ymax=x+se))
gp +geom_line(aes(linetype=group), size = .4) + 
    geom_point(aes(shape=group)) + 
    geom_ribbon(alpha = 0.15, linetype=0) + 
    #geom_errorbar(aes(ymax=x+se, ymin=x-se)) +
    ylim(0, 1) + 

jvcasillas

#####################################
# Functions for linguistic research #
# Joseph V. Casillas                #
# Last update: 12-10-2014           #
#####################################

# Calculate d prime in discrimination experiments. 
# Takes a df with the columns:
#    - stimDiff (actual stimuli are different)
#    - stimSame (actual stimuli are the same)

jvcasillas

euc.dist <- function(x1, x2){
  x <- sqrt(sum((x1 - x2)^2))
  return(x)
}

jvcasillas

# Plot vowel formants
# - Include elipses, 
# - group/vowel means
# - greyscale

# Load tidyverse and phonR package (I choose this because you said you 
# had alrady used it)
library(phonR)
library(tidyverse)

jvcasillas

# Load libraries
library("tidyverse")
library("TOSTER")

# Set seed for reproducibility
set.seed(10)

# Working memory means for 3 groups
wm_n   <- rnorm(mean = 8.67, sd = 1.15, n = 12)
wm_nin <- rnorm(mean = 9.04, sd = 2.11, n = 26)