The chapter provides a background and overview of R. A nice YouTube video can introduce you to both: Getting Started with R and R Studio
The setup instructions could be helpful if you have not done so already.
Prior to pulling in data, you’ll need to make sure that you’re “pointing” R in the right direction to find the file. To figure where R is pointing send this or, more realistically, type this in the console:
getwd()
The file path is where R will look to get your file. If you are working in a R Project, which you should always be doing, it will go to the main folder for that project. If you have a subfolder in that R Project then you can direct the import to that folder as noted below in the first, *.csv, example. You can, although you really shouldn’t have to, set your file path using by using the command setwd(FILE PATH) or using the dropdown menu in R Studion under Session.
Go the source and check out the RStudio Data Import Cheatsheet.
Below are commonly used functions for importing data in several CLU classes. A more exhaustive list can be found in Wickham’s R4DS.
Step 1. Load Libaries
The functions we use rely on the following three packages all of which are contained within the tidyverse. However, you could load them seperately such as:
# Load Libraries for Import
library(readr) # for *.csv and *.dat in class
library(haven) # for *.sav (spss files) and *.dta (stata files)
library(readxl) # for *.xlsx and *.xls files
Or more simply as a collective pacakage:
library(tidyverse)
Step 2. Import the Data: Examples
data.tb <- read_csv("dsur.csv")
data.tb <- read_csv("03-data/data.csv")
data.tb <- read.csv("data.csv", header=T)
data.tb <- read_excel("data.xlsx")
data1.tb <- read_excel("data.xlsx", sheet = "Name" or #) # if you have multiple sheets
haven:
data.tb <- read_sav("data.sav")
haven but loading inline:
data.tb <- read_sav("data.sav")
data.tb <- read_delim("data.dat", delim="\t")
data.tb <- read.delim("data.dat") # non Tidyverse way
haven:
data.tb <- haven::read_dta("data.dta")
Notice:
All these require that R is going to the correct working directory (i.e., where your data lives). If you’re working in an R Project then you’re good to go unless you have a subfolder like in the second example where we direct it to the subfolder.
You can use the :: to designate the library to be loaded. In this case you don’t have to do library() in an earlier chunk. An example can be seen in #9 where the haven library is loaded inline.
GitHub is a great resource for data analysis projects. You can certainly download the data and then put it in your R Project folder. You can also directly import from the web. The following is an example of how to import the data directly form a website.
The data here is from the crowdsourcing study on soccer red cards. Accessible introductions to the data can be found in Nature and in Andrew Gelman’s blog.
A GitHub user has put the data up using a .csv file. Here’s the URL to the user’s account. However, this is not the URL you need to import. Click on “View Raw” and the following URL will appear along with some unsightly data ‘https://raw.githubusercontent.com/cgonza12/Soccer/master/data.csv’. This is your link to get the data. Follow the steps
A) Install and load the required library.
library(RCurl) # load the library.
B) Paste the URL into the "" space.
redcard <-read.csv(text=getURL("https://raw.githubusercontent.com/cgonza12/Soccer/master/data.csv"))
C. Make it a tibble for accessibility
redcard.tb <- tibble::as.tibble(redcard)
redcard.tb
A tibble: 146,028 x 27
playerShort player club leagueCountry birthday height weight position games
<fct> <fct> <fct> <fct> <fct> <int> <int> <fct> <int>
1 lucas-wilc… Lucas… Real… Spain 31.08.1… 177 72 Attacki… 1
2 john-utaka John … Mont… France 08.01.1… 179 82 Right W… 1
3 abdon-prats " Abd… RCD … Spain 17.12.1… 181 79 NA 1
4 pablo-mari " Pab… RCD … Spain 31.08.1… 191 87 Center … 1
5 ruben-pena " Rub… Real… Spain 18.07.1… 172 70 Right M… 1
6 aaron-hugh… Aaron… Fulh… England 08.11.1… 182 71 Center … 1
7 aleksandar… Aleks… Manc… England 10.11.1… 187 80 Left Fu… 1
8 alexander-… Alexa… Norw… England 04.04.1… 180 68 Defensi… 1
9 anders-lin… Ander… Manc… England 13.04.1… 193 80 Goalkee… 1
10 andreas-be… Andre… 1899… Germany 13.03.1… 180 70 Right F… 1
... with 146,018 more rows, and 18 more variables: victories <int>, ties <int>,
defeats <int>, goals <int>, yellowCards <int>, yellowReds <int>,
redCards <int>, photoID <fct>, rater1 <int>, rater2 <int>, refNum <int>,
refCountry <int>, meanIAT <dbl>, nIAT <int>, seIAT <dbl>, meanExp <dbl>,
nExp <int>, seExp <dbl>
You’ll use OSF throughout your time at CLU. Similar to GitHub, OSF is a great place to find projects and share material. Just like GitHub, you could easily download the data and place it in an R Project. In this example, however, I’ll provide code showing how you can import directly from OSF. Here we’ll get data used to conduct the P-curve analysis for the Power Poses discussed at Data Colada by the authors and also in Slate article by Andrew Gelman.
A) httr package is need for the GET command. Install and load.
library(httr)
B) Here you need: 1) The name of the file and in this case it’s an .xlsx file. 2) The URL followed by /?action=download and saved as data 3) Our file is in data and we’ll call it pp.xlsx to preserve the original extension. The names data and pp are irrelevant.
if (!file.exists('p-curve disclosure table for Power Posing review - 2016 09 26.xlsx')) {
data <- GET('osf.io/2fq9c//?action=download',
write_disk('pp.xlsx', overwrite = TRUE))
}
C) Using the readxl package you can open the saved file.
pp.tb <- readxl::read_excel('pp.xlsx')
pp.tb
A tibble: 33 x 13
Study `Categorize d.v… `dependent vari… `Quoted Predict…
<chr> <chr> <chr> <chr>
1 Alle… 3 number of pretz… We hypothesized…
2 Arne… 3 Position of cha… Posture, as it …
3 Bohn… 3 pain control Experiment 1 te…
4 Bohn… 3 pain control Specifically, w…
5 Brin… 2 self-evalaution In this experim…
6 Carn… 3 hormones and ri… We hypothesized…
7 Carn… 3 hormones and ri… We hypothesized…
8 Cesa… 3 gambling If embodied eff…
9 Cesa… 3 gambling Study 2 manipul…
10 Cudd… 3 Evaluated perfo… Specifically, t…
... with 23 more rows, and 9 more variables: `Study
Design` <chr>, `Key Statistical Result` <chr>, `Quoted
Results` <chr>, Results <chr>, `Quoted Robustness
Results` <chr>, `Robustness Results` <chr>, X__1 <chr>,
`Quoted Feelings Of Power` <chr>, `Feelings Of Power
Results` <chr>