Data Visualization for Political Scientists

class: center, middle, inverse, title-slide

# Data Visualization for Political Scientists
## Workshop at the University of Cologne
<html>
<div style="float:left">

</div>
<hr color='#EB811B' size=1px width=796px>
</html>
### Felix Haass
### 15 Januar 2018

---

# Introduction - Me

**Felix Haass**

<i class="fa fa-university"></i> ABI Freiburg / GIGA Hamburg

<i class="fa fa-search"></i> Political economy; conflict, democratization, United Nations

<i class="fa fa-twitter"></i> [felixhaass](http://twitter.com/felixhaass)

<i class="fa fa-globe"></i> [http://felixhaass.github.io](http://felixhaass.github.io)

<i class="fa fa-envelope-o"></i> [haass@giga-hamburg.de](mailto:haass@giga-hamburg.de)

---

# Introduction - You

Briefly introduce yourself!

<i class="fa fa-question"></i> Your Name

<i class="fa fa-search"></i> Your Research

<i class="fa fa-binoculars"></i> Your Motivation for Participation in the Workshop

---

# Learning Objectives

- Learn principles of how to use the `ggplot2` package to create informative data visualizations  
- We will learn mechanics of the software package, rather than good visualization principles. For infos on that, see course page (watch out for the dataviz ninja, though!).   
- Know where to find help

---

# Course Principles

- Ask questions at any time!
- Collaborate with your neighbours! 
- Individual preferences/data visualization problems are most welcome!

<div align = "center">
<blockquote class="twitter-tweet" data-lang="de">
<p lang="en" dir="ltr">You&#39;re doing it right if you get frustrated: if you&#39;re not frustrated, you&#39;re (probably) not stretching yourself mentally</p>&mdash; Hadley Wickham (@hadleywickham) 
<a href="https://twitter.com/hadleywickham/status/565516733516349441?ref_src=twsrc%5Etfw">11. Februar 2015</a>
</blockquote>
<script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
</div>

---

# Outline of the Course

| Time |	Topic |	
|-------|-------|
| 9.30-10.00 | Session 1: Introduction and R refresher  |
| 10.00-10.30 | Session 1 continued: Reading data, project organization |	
| *10.30-10.45* |	*Coffee Break*	| 
| 10.45-12.15	| Session 2: The logic of a grammar of graphics & its implementation in ggplot2 |	
| *12.15-13.15* |	*Lunch Break*	| 
|13.15-14.45 |	Session 3: Use-Cases I - Facets and small multiples; sorting facets |
| *14.45-15.00* |	*Coffee Break*	|
| 15.00-16.00 |	Session 4: Use-Case II - Coefficient plots |	
| 16.00-17.00 |	Session 5: Wrap-up - Exporting plots; questions; where to get help |

---

# R Refresher - What is R?

- `R` is a programming language for statistical analysis
    - "Programming language": repeat tedious tasks; replicability; connect to unusual data sources (Twitter, web sites, text documents, ...)
    - "statistical analysis": many packages to do any statistical analysis you want
- `RStudio` is the interactive software with which we write and execute R code, plot things, view the R memory environment (...and much more)

---

# R Refresher - Libraries

R uses different **libraries** or **packages** to load specific functions (read excel files, talk to Twitter, generate plots, ...): https://cran.r-project.org/. You load a package or a library with the command

```r
library(read_excel) # read_excel is the package name (without quotation marks)
```

If a command throws an error, chances are you either

- forgot to load the respective library
- have a syntax error - **R is case sensitive!**

To install a **package** we use:

```r
install.packages("gapminder") # with quotation marks!
```

---

# R Refresher - Assignment

In R, we assign **stuff** (numbers, characters, data frames) to **things** (objects)

```r
url <- "http://gmi.bicc.de/index.php?page=ranking-table"
```

- `url`: object, in this case: a character vector 
- `"http://gmi.bicc.de/index.php?page=ranking-table"`: "stuff" (URL, could be any text or number)   
- `<-`: assign command, type `<` and `-` (shortcut: <kbd>alt</kbd> + <kbd>-</kbd> in RStudio)

---

# R Refresher - Objects

In R, everything is an object--and you can have multiple objects in your memory at the same time!

```r
# 1st object: assign numbers to a vector
numbers <- 1:5

# 2nd object: read data from an excel sheet
sipri <- read_excel("./data/SIPRI-Milex-data-1949-2016_cleaned.xlsx", 
                    sheet = 5, 
                    na = c("xxx", ". ."))
```

Executing this command yields to objects in memory, `numbers` the vector of numbers and the data frame `sipri`.

---

# R Refresher - Data Frames

Data frames are rectangular data tables, like an Excel spreadsheet.

```r
library(gapminder)
library(tidyverse)

gapminder
```

```
## # A tibble: 1,704 x 6
##    country     continent  year lifeExp      pop gdpPercap
##    <fctr>      <fctr>    <int>   <dbl>    <int>     <dbl>
##  1 Afghanistan Asia       1952    28.8  8425333       779
##  2 Afghanistan Asia       1957    30.3  9240934       821
##  3 Afghanistan Asia       1962    32.0 10267083       853
##  4 Afghanistan Asia       1967    34.0 11537966       836
##  5 Afghanistan Asia       1972    36.1 13079460       740
##  6 Afghanistan Asia       1977    38.4 14880372       786
##  7 Afghanistan Asia       1982    39.9 12881816       978
##  8 Afghanistan Asia       1987    40.8 13867957       852
##  9 Afghanistan Asia       1992    41.7 16317921       649
## 10 Afghanistan Asia       1997    41.8 22227415       635
## # ... with 1,694 more rows
```

---

# Review: R Refresher

- **Libraries** are your friend! `library()` (load) them or `install.packages()` them!
- **Assign** stuff. Use `<-` for assignments!
- **Multiple objects** can and should exist in memory (if you lose track of the objects you're juggling with, check the environemnt panel at the upper right of RStudio)
- **Data frames** are the most important type of object. Don't stop until you have your data in a data frame (check data type with `class()`)!
- Bonus tip: use the help function `help(command_name)` if you can't remember a command's options.

---

# Organizing your Code

Having a structured way to organize your R code is useful for reproducibility (and your future sanity!)

There are two ways to improve your R code organization:

1. Folder Structure
2. RStudio Projects

---

# Organizing your Code: Folder Structure

A useful way to organize your project folders:

```bash

project_name/       # name of your project
|-- code/           # here go all the .R script files
|-- data/           # here's your data
|   |-- input/       # raw input data file (experimental results, existing datasets)
    |-- output/      # transformed and cleaned datasets for analysis
|-- manuscript/     # your manuscript, i.e. .docx or LaTeX files
|-- figures/        # your figures as separate files 
|-- output/         # tables

```

---

# Organizing your Code: RStudio Projects

An RStudio project takes care of several useful steps in your project. When you load an RStudio project, the following steps are taken:

- Sets the working directory to the project directory
- A new R session (process) is started (with no libraries)
- Previously edited source documents are restored into editor tabs
- Other RStudio settings (e.g. active tabs, splitter positions, etc.) are restored to where they were the last time the project was closed.
- [...and others, see https://support.rstudio.com/hc/en-us/articles/200526207-Using-Projects]

---

# Organizing your Code: Setting up an RStudio Project {.columns-2}

In RStudio, go to File => New Project => "Existing Directory"

---

# Reading Data

To read `.csv` files, the the `read_csv()` function in the `readr` package is useful (automatically loaded through `library(tidyverse)`).

To read Excel files, use the `read_excel()` function from the `readxl` package, which needs to be loaded separately.

To read files from Stata or SPSS, use `read_dta()` or `read_spss()` from the `haven` package, which needs to be loaded separately.

Example:

```r
library(tidyverse)
library(readxl)

sipri <- read_excel("./data/SIPRI-Milex-data-1949-2016_cleaned.xlsx", 
                    sheet = 5, na = c("xxx", ". ."))
```

To read R files (`.rda` or `.rdata`), simply use `load("name_of_my_file.rda"`)

Also useful: the `rio` package!

---

# Exercise

1. Create organized folder structure
2. Download the SIPRI data from the course website
3. Create an R script to read the data (hint: remember that the Excel file has several sheets!)

---

# Solution

```r
library(tidyverse)
library(readxl)
```

```
## Warning: package 'readxl' was built under R version 3.4.3
```

```r
sipri <- read_excel("./data/SIPRI-Milex-data-1949-2016_cleaned.xlsx", 
                    sheet = 5, na = c("xxx", ". ."))

sipri_plot <- sipri %>%
  # from wide to long format with the `gather function
  gather(key = year, 
         value = military_expenditure, 
         -Country) %>% 
  ggplot(., aes(x = year, 
                y = military_expenditure, 
                group = Country)) + 
  geom_line(alpha = 0.5)
```

---

# Solution

```r
print(sipri_plot)
```

![](session1_intro_files/figure-html/unnamed-chunk-6-1.png)