2014-2015 presentation series: Switch2R

• 

  • is a free general purpose statistical software and programming language
  • is the lingua franca of statisticians
  • is eating the lunch of SAS and SPSS
  • learning is hard, so we’re going to help each other and form a local R users group

• Narragansett Bay Area R Users
• http://www.meetup.com/Narragansett-Bay-Area-R-Users/
• BRAUN kick-off will be the second Switch2R/AMPHIBIAN meeting (November)

• Why

  • R User Groups are supportive environments
  • The closest R Users Group is in Boston
  • There has got to be some experienced R users in the area who can help us

Switch2R Series

October 2014: R Basics

• Install R
• Install and load packages

• Use R to

  • read in a data set and
  • generate some descriptive statistics and
  • basic plots
• R environments

• Learning resources for R

November 2014: Clean Data with R

• Use R to clean data

  • make new variables
  • label variables and values
  • produce dataset documentation

Monday, November 10, 2014, 2-3pm

December 2014 Analysis with R

• Use R to do multivariable analysis

  • Linear regression
  • Logistic regression
  • Survival analysis
  • Repeated measures mixed models

Monday, December 8, 2014, 2-3pm.

January 2015 Graphics with R (1/2)

Use R to produce high-quality, publication quality statistical graphics

• Two systems

  • base graphics (http://goo.gl/e51V0a)
  • lattice (http://goo.gl/3EnSfa)

Monday, January 12, 2015, 2-3pm

February 2015: Graphics with R (2/2)

Use R to produce high-quality, publication quality statistical graphics

• Two more systems

  • ggplot2 (http://goo.gl/Ri05Pw)
  • ggvis (http://ggvis.rstudio.com/)

Monday, February 9, 2015, 2-3pm.

March 2015 Reproducible research with R

• R/Markdown

  • Documents
  • Slides

Monday, March 16, 2015, 2-3pm.

April 2015: GIS with R

Maps and geospatial data analysis

Monday, April 13, 2015, 2-3pm.

May 2015: Power and sample size

• Canned programs and
• Simulation approach

Monday, May 11, 2015, 2-3pm.

June, July Aug. 2015: Summer break

No meetings July 2015 Latent Variable Models Workshop

• Longitudinal Data Analysis Using R/lavaan and Mplus
• Weds-Fri July 29 - 31, 9:30am - 4:00pm

What is missing?

• EMA data
• see (http://www.rdocumentation.org/packages/funreg)
• Something else?

Installing R

Installing R

Install R by googling for “R Cran”

Installing R Packages

What are R Packages

• R has base packages and add-on packages

• R base packages load automatically

• R Users write packages and post them on the web (repositories)

  • CRAN
  • Bioconductor
  • Github

Installing R & Packages

Here I am installing Rcmdr

Read in Data

ASCII Data

Example at UCLA Institute for Digital Research and Education

(http://www.ats.ucla.edu/stat/paperexamples/atkins_mlm/default.htm)

Atkins, D. C. (2005). Using multilevel models to analyze couple and family treatment data: Basic and advanced issues. Journal of Family Psychology, 19, 98-110.

post-print version (PDF) (http://goo.gl/7uZiEH)

psycnet (http://psycnet.apa.org/journals/fam/19/1/98/)

findit at Brown full text (http://goo.gl/D6m2ID)

# Read data from UCLA Web site
url <- "http://www.ats.ucla.edu/stat/paperexamples/atkins_mlm/Atkins_JFP_data.txt"
data <- read.csv(url, sep="\t", header=TRUE)
# show the first six lines
head(data)

##   ID SEX THERAPY TIME      DAS PILOT MISS M.IND
## 1  1   0    -0.5    0 94.51204     1    0     1
## 2  1   0    -0.5   13 87.53364     1    0     1
## 3  1   0    -0.5   26 81.46659     1    1     1
## 4  1   0    -0.5   35 83.44614     1    1     1
## 5  1   1    -0.5    0 81.27981     1    0     1
## 6  1   1    -0.5   13 68.80343     1    0     1

Other ways to read in data

• R Package: foreign

  • SPSS
  • SAS
  • Stata
  • and others

• R package: sjPlot

  • specifically for SPSS

Basic descriptive statitics

summary(data)

##        ID             SEX         THERAPY          TIME      
##  Min.   :  1.0   Min.   :0.0   Min.   :-0.5   Min.   : 0.00  
##  1st Qu.: 34.0   1st Qu.:0.0   1st Qu.:-0.5   1st Qu.: 9.75  
##  Median : 67.5   Median :0.5   Median : 0.0   Median :19.50  
##  Mean   : 67.5   Mean   :0.5   Mean   : 0.0   Mean   :18.50  
##  3rd Qu.:101.0   3rd Qu.:1.0   3rd Qu.: 0.5   3rd Qu.:28.25  
##  Max.   :134.0   Max.   :1.0   Max.   : 0.5   Max.   :35.00  
##       DAS             PILOT             MISS             M.IND       
##  Min.   : 40.66   Min.   :0.0000   Min.   :0.00000   Min.   :0.0000  
##  1st Qu.: 77.85   1st Qu.:0.0000   1st Qu.:0.00000   1st Qu.:0.0000  
##  Median : 88.19   Median :0.0000   Median :0.00000   Median :0.0000  
##  Mean   : 89.00   Mean   :0.2239   Mean   :0.07836   Mean   :0.1567  
##  3rd Qu.: 99.12   3rd Qu.:0.0000   3rd Qu.:0.00000   3rd Qu.:0.0000  
##  Max.   :154.31   Max.   :1.0000   Max.   :1.00000   Max.   :1.0000

More descriptive statistics

library(pipeR)
library(xtable)
library(plyr)

data %>>%
  {do.call(data.frame, 
           list(mean = apply(., 2, mean),
                sd = apply(., 2, sd),
                median = apply(., 2, median),
                min = apply(., 2, min),
                max = apply(., 2, max),
                n = apply(., 2, length)))} %>>%
  xtable(caption = "Summary of the data", 
         digits=2) %>>%
  print(type = "html", caption.placement="top",
        html.table.attributes = FALSE)

R Environments

R Environments

• Regular R (Revolution R Open)
• RStudio
• RCmdr

Regular R (Revolution R Open looks and feels the same)

RStudio

RCmdr

Where to get more information

Training

• COURSERA Data Science from Johns Hopkins University
  • The Data Scientist’s Toolbox
  • R Programming
  • Getting and Cleaning Data
  • Exploratory Data Analysis
  • Reproducible Research
  • Statistical Inference
  • Regression Models
  • Practical Machine Learning
  • Developing Data Products

Web Sites

• R-Bloggers (http://www.r-bloggers.com/)
• 2 minute R (http://www.twotorials.com/)
• Quick-R (http://www.statmethods.net/)
• Search all R documentation (http://www.rdocumentation.org/)
• Google (http://www.google.com/)
• Twitter (#rstats)

See you next time

Use R to clean data

Monday, November 10, 2014, 2-3pm

Inaugural BRAUN meeting