RonR! First seminar

Introduction to R and its strengths

Jorge Cimentada
5th of March of 2016

What is R?

R is a language and environment for statistical computing and graphics
An effective data handling and storage facility;
A large, coherent, integrated collection of intermediate tools for data analysis;
Graphical facilities for data analysis and display either on-screen or on hardcopy;
Simple and effective programming language which includes conditionals, loops, user-defined recursive functions and input and output facilities.

Do you know what General Public License or GNU is?

You have the freedom to run the program as you wish, for any purpose (freedom 0).
You have the freedom to study how the program works, and change it so it does your computing as you wish (freedom 1).
You have the freedom to redistribute copies so you can help your neighbor (freedom 2).
You have the freedom to distribute copies of your modified versions to others (freedom 3).

How does R work?

Getting started

Go to www.r-project.org
Downloads: CRAN
Set your Mirror: Anywhere in Spain is fine.
Select 'Download R for Windows'.
Select base.
Select R 3.2.3 for Windows
- The others are if you are a developer and wish to change the source code.

Let's go to R!

R interface

Downloading Rstudio

How does Rstudio look?

R studio interface

What are R's strengths?

Adapted from 'R for Stata users' by Muenchen and Hilbe

Comprehensible repertoire of statistical techniques
R is directly accessible from many statistics packages like SAS, SPSS and Stata.
Every aspect of R is open for anyone to modify in any way they like.
R also offers the very flexible and powerful Grammar of Graphics approach. Developers have even gone so far as replacing R's core graphical system.
R is free. This helps to attract developers and is a major reason that there are so many add-on packages for it.

R graphics

par(mfrow = c(1, 2))
plot(iris$Sepal.Length, iris$Petal.Length, col = iris$Species, main = "Sepal vs Petal Length in Iris")
plot(iris$Sepal.Length, iris$Sepal.Width, col = iris$Species, main = "Sepal Length vs Width in Iris")

plot of chunk unnamed-chunk-1

R graphics

Problem = Explanatory graphs for teaching

plot of chunk unnamed-chunk-2

R graphics

*Problem = Showing overlapping variation using Venn diagrams plot of chunk unnamed-chunk-3

R graphics

Problem = Showing relationships between data points This is among the most recognized graphs in R!

R graphics - ggplot2 package

Problem = Nicely formatted barplot with minimum code

library(ggplot2)
ggplot(diamonds, aes(clarity, fill=cut)) + geom_bar()

plot of chunk unnamed-chunk-4

R graphics - ggplot2 package

Problem = Visualizing a bivariate relationship filtered by two variables

library(ggplot2)
qplot(wt, mpg, data=mtcars, size=qsec, color=factor(carb))

plot of chunk unnamed-chunk-5

R graphics - ggplot2 package

Problem = Visualizing a bivariate relationship by a categorical variable

library(ggplot2)
qplot(mpg, wt, data = mtcars, facets = cyl ~ ., geom = c("point", "smooth")) + 
    coord_flip()

plot of chunk unnamed-chunk-6

R graphics - ggplot2 package

_Problem = Combine existing graphs in a nice format _* **Note: There's some other code going on in the back! Not that easy.

multiplot(p1, p2, p3, p4, cols = 2)

plot of chunk unnamed-chunk-8

Multilevel PSM

_Problem = Visualizing complex statistical techniques http://jason.bryer.org/multilevelPSA/

Psychology Reproducibility

_Problem = Plotting distributions next to scatterplots

Country performance in PISA

Problem = Making nice graphical summaries

Likert package

Problem = Visualize Likert type questions quickly likert

Interactive graphs

_Problem = Exploring your data visually in a fast way _ ** Run this in your R console **

library(openintro)
edaPlot(mtcars)

Interactive graphs

Problem = Getting QUALITY visualizations with almost no work.

Let's go to the console to get to know ggvis

Advantages of ggvis over other graph packages:

HTML and Javascript graphics are the ones actually doing the work behind ggvis.

For more info, go to ggvis

Interactive graphs

_Problem = Getting QUALITY visualizations exportable to websites. _ ** Run this on your R console **

library(plotly)
set.seed(100)
d <- diamonds[sample(nrow(diamonds), 1000), ]
plot_ly(d, x = carat, y = price, text = paste("Clarity: ", clarity),
        mode = "markers", color = carat, size = carat)

For more see plotly

Interactive graphs

Shiny has also interactive graphs which can be included in websites
Simpler example

Interactive graphs

Let's play around!

library(ggvis)
span <- waggle(0.2,1)
span <- left_right(0.2,1,step = 0.1)

mtcars %>%
    ggvis(~mpg, ~disp) %>%
    layer_lines() %>%
    layer_smooths(span=span)

The first 'span' variable will adjust the smoothness automatically You can adjust the smoothness with your left-right arrow with the second span

R graphics

rCharts <!– MotionChart generated in R 3.2.4 by googleVis 0.5.10 package –> <!– Wed Mar 16 10:42:11 2016 –>

R maps

plot of chunk unnamed-chunk-13

R maps

Crime density in Baltimore

R maps

Summary statistics ouput

** Run this in the console **

library(stargazer)
stargazer(mtcars[c("mpg","hp","drat")], type = "text",
 title="Descriptive statistics/selected variables", digits=1, out="table2.txt", flip=TRUE,
 covariate.labels=c("Miles/(US)gallon","Gross horsepower","Rear axle ratio"))

Regression statistics ouput

** Run this in the console **

library(stargazer)
mtcars$fast <- as.numeric((mtcars$mpg > 20.1))

m1 <- lm(mpg ~ hp + drat + factor(gear), data=mtcars)
m2 <- glm(fast ~ hp + drat + am, family=binomial(link="logit"), data=mtcars)

stargazer(m1, m2,type="text", dep.var.labels=c("Miles/(US) gallon","Fast car (=1)"), covariate.labels=c("Gross horsepower","Rear axle ratio","Four foward gears","Five forward gears","Type of transmission (manual=1)"))

Where did you think this presentation was done?

Checkout Rmarkdown and also here

But of course, it's not built to write research papers, it's more for:

Blog posts;
Posts to StackExchange sites;
Materials for training workshops;
Short consulting reports;
Exploratory analyses as part of a larger project;

There's also slidify

You can also save it as pdf

Exercises for you students

With the swirl package you can program exercises and homeworks that students can do anywhere.

Try it out yourself:

library("swirl")
swirl()