gkane26 / rblinks

R package to import data from Med Associates operant chamber blinks task

rblinks: R package to read MedPC Blinks task data.

This package contains 3 functions:

• med_to_dt: reads a MedPC raw data file to an R data table
• save_blinks_data: saves file to a specified path
• load_blinks_data: loads data from a specified file

Installation

First, make sure you have the devtools package installed. If not, run:
install.packages("devtools")

Next, install rblinks directly from github:
devtools::install_github("gkane26/rblinks")

Instructions for use

First, get the path to MedPC data files. These can be typed in manually, or selected via a GUI in a number of ways. I think the tcltk package is a good option across platforms. If using a mac, this will require XQuartz is installed. Example:
files = tcltk::tk_choose.files()

Next, use the med_to_dt function to get a data frame of trial-by-trial data:

library(rblinks)
blinks_dat = med_to_dt(files[1])

Data can be saved to either a csv or Rdata file using the save_blinks_data function:

save_file_name = save_blinks_data(blinks_dat, path="~/Desktop")

Previously saved data can be loaded using the load_blinks_data function. Multiple files can be loaded into a single data table by passing a vector of file names:

blinks_dat = load_blinks_data(save_file_name)
blinks_dat = load_blinks_data(c(save_file_name, save_file_name))

To see additional options or for further details, see documentation. For example: ?med_to_dt

Further Analysis

At the moment, this package does not implement any further statistical analysis. Since the package uses data.table, further analysis is easy and convenient to do (see Introduction to data.table). After loading a data table containing data from multiple groups of subjects and multiple experimental conditions, accuracy and reaction time can be found:

blinks_data = rblinks::load_blinks_data(file=vector_of_data_files)
accuracy_data = blinks_data[, accuracy := mean(correctChoice), .(subject, group, condition, pCorrectBlink)]
rt_data = blinks_data[, RT := mean(RT), .(subject, group, condition, pCorrectBlink)]

These variables can be easily plotted using ggplot2:

library(ggplot2)
library(Hmisc)

accuracy_plot = ggplot(accuracy_data, aes(x=pCorrectBlink, y=accuracy, color=condition)) +
    facet_wrap(~group)
    stat_summary(fun,y=mean, geom="point", size=2) +
    stat_summary(fun.y=mean, geom="line") +
    stat_summary(fun.data=mean_se, geom="errorbar", width=.03) +
    theme_classic()
accuracy_plot