Intro to data visualisation

class: center, middle, inverse, title-slide

# Intro to data visualisation
### Jakub Nowosad <br><a href="mailto:nowosad.jakub@gmail.com">nowosad.jakub@gmail.com</a>
### 2017-04-26

---

## Data visualisation in R

- R has several systems for making graphics - **base R**, the **lattice** package, and the **ggplot2** package are among the most popular ones

![](Intro_to_data_visualisation_files/figure-html/unnamed-chunk-1-1.png)![](Intro_to_data_visualisation_files/figure-html/unnamed-chunk-1-2.png)![](Intro_to_data_visualisation_files/figure-html/unnamed-chunk-1-3.png)

---
## **ggplot2** - basic information

- **ggplot2** is an implementation of the grammar of graphics (["Grammar of Graphics"](https://www.amazon.com/Grammar-Graphics-Statistics-Computing/dp/0387245448) book written by Leland Wilkinson)
- data frames are (most of the times) an input data 
- the **ggplot2** package has two functions for plot creation - `qplot()` and `ggplot()`
- **ggplot2** documentation is available at http://docs.ggplot2.org

```r
install.packages("ggplot2")
```

```r
library('ggplot2')
```

---
## **ggplot2** - the first plot

```r
df <- data.frame(A = c(1:10), B = seq(22, 4, by= -2))
ggplot(data = df, aes(x = A, y = B)) + geom_point()
```

---
## ggplot2 - basic dictionary

- `geom` - type of a plot ("histogram", "boxplot", "point", etc.)
- aesthetics (`aes`) - visual properties of geoms, such as color, size, shape, ...
- faceting (`facet`) - panels with different subsets of the data

---
## Dataset

```r
install.packages('gapminder')
library('gapminder')
```

```r
gapminder <- readRDS('data/gapminder.rds')
```

http://www.gapminder.org/data/

http://github.com/jennybc/gapminder

http://www.youtube.com/watch?v=jbkSRLYSojo

```r
gapminder2007 <- subset(gapminder, year==2007)
```

---
## Histogram

```r
ggplot(data=gapminder2007, aes(x=gdpPercap)) + 
        geom_histogram()
```

---
## Histogram

.pull-left[

```r
ggplot(data=gapminder2007, aes(x=gdpPercap)) + 
        geom_histogram(binwidth=10000)
```

<img src="Intro_to_data_visualisation_files/figure-html/unnamed-chunk-9-1.png" width="400px" />
]

.pull-right[

```r
ggplot(data=gapminder2007, aes(x=gdpPercap)) + 
        geom_histogram(binwidth=800)
```

<img src="Intro_to_data_visualisation_files/figure-html/unnamed-chunk-10-1.png" width="400px" />
]

---
## Bar plot

There are two types of bar plots in **ggplot2**:
- `geom_bar()` makes the height of the bar proportional to the number of cases in each group
- `geom_col()` makes the height of the bars to represent values in the data

```r
ggplot(data=gapminder2007, aes(x=continent)) + geom_bar()
```

---
## Bar plot

```r
library('dplyr')
gapminder2007_2 <- gapminder2007 %>% group_by(continent) %>% summarise(number=n())
ggplot(data=gapminder2007_2, aes(x=continent, y=number)) + geom_col()
```

---
## Line plot

```r
gapminder2 <- gapminder %>% group_by(year) %>% summarise(mean.lifeExp=mean(lifeExp, na.rm=TRUE))
ggplot(data=gapminder2, aes(x=year, y=mean.lifeExp)) + geom_line()
```

---
## Scatter plot

```r
ggplot(data=gapminder2007, aes(x=gdpPercap, y=lifeExp)) + geom_point()
```

---
## Box plot

```r
ggplot(data=gapminder2007, aes(x=continent, y=lifeExp)) +
        geom_boxplot()
```

---
## Aesthetic attributes for quantitative variables (1)

```r
ggplot(data=gapminder2007, aes(x=gdpPercap, y=lifeExp, color=pop)) +
        geom_point()
```

---
## Aesthetic attributes for quantitative variables (2)

```r
ggplot(data=gapminder2007, aes(x=gdpPercap, y=lifeExp, size=pop)) +
        geom_point()
```

---
## Aesthetic attributes for qualitative variables (1)

```r
ggplot(data=gapminder2007, aes(x=gdpPercap, y=lifeExp, color=continent)) +
        geom_point()
```

---
## Aesthetic attributes for qualitative variables (2)

```r
ggplot(data=gapminder2007, aes(x=gdpPercap, y=lifeExp, shape=continent)) + 
        geom_point() + scale_shape(solid = FALSE)
```

---
## Scales

```r
ggplot(data=gapminder2007, aes(x=gdpPercap, y=lifeExp, color=pop)) +
        geom_point()+ 
        scale_colour_gradientn(colours=rainbow(5))
```

---
## Data transformations

```r
ggplot(data=gapminder2007, aes(x=gdpPercap, y=lifeExp, color=log10(pop))) +
        geom_point()
```

---
## Facets

- Two functions, `facet_grid()` and `facet_wrap()` can be used to plot the subsets of data together

```r
ggplot(data=gapminder2007, aes(x=lifeExp)) + 
        geom_histogram() + 
        facet_wrap(~continent)
```

---
## Plot customizing - labs

```r
p <- ggplot(data=gapminder2007, aes(x=lifeExp, y=gdpPercap, color=continent)) +
        geom_point()
p <- p + labs(x='Life expectancy', y=NULL,
              title='GDP vs Life expectancy', color='Continent: ')
p
```

---
## Plot customizing - themes

```r
p <- p + theme_bw()
p
```

---
## Saving plots

```r
p <- ggplot(data=gapminder2007, aes(x=lifeExp, y=gdpPercap, color=continent)) +
        geom_point() +
        labs(x='Life expectancy', y=NULL, title='GDP vs Life expectancy', color='Continent: ') +
        theme_bw()
p
```

```r
ggsave(filename = "Plot.pdf", plot = p)
```

```r
ggsave(filename = "Plot.png", plot = p, dpi = 300)
```

---
## That's not all (folks)

- More geoms:
   - `geom_density()`
   - `geom_jitter()`
   - `geom_text()`
   - ... 
- Axes and legends customizing
- Additional packages (**ggplot2** extensions) - http://www.ggplot2-exts.org/
- Spatial data visualisation
- ... and many more

---
## Interactive plots

- R has many interfaces to interactive visualization libraries, such as `plotly`, `dygraphs`, `rCharts`, `googleVis`, etc.
- What's even better is that R allows for building interactive web applications using the `shiny` package

---
## `plotly`

```r
# devtools::install_github("ropensci/plotly")
library('plotly')
ggplotly(p=p)
```

<div id="htmlwidget-e836405ebb8808948a92" style="width:100%;height:400px;" class="widgetframe html-widget"></div>


<script type="application/json" data-for="htmlwidget-e836405ebb8808948a92">{"x":{"url":"widget_unnamed-chunk-30.html","options":{"xdomain":"*","allowfullscreen":false}},"evals":[],"jsHooks":[]}</script>

---
## `dygraphs`

```r
library('dygraphs')
library('tidyr')
gapminder3 <- gapminder %>%
        group_by(year, continent) %>%
        summarise(mean.lifeExp=mean(lifeExp, na.rm=TRUE)) %>% 
        spread(continent, mean.lifeExp)
dygraph(gapminder3) %>% dyRangeSelector()
```

<div id="htmlwidget-fca20722a6512a86b2b4" style="width:95%;height:300px;" class="widgetframe html-widget"></div>


<script type="application/json" data-for="htmlwidget-fca20722a6512a86b2b4">{"x":{"url":"widget_unnamed-chunk-32.html","options":{"xdomain":"*","allowfullscreen":false}},"evals":[],"jsHooks":[]}</script>

---
## Resources:

- [ggplot2](http://ggplot2.tidyverse.org/) - an official website of the `ggplot2` package
- [An introduction to ggplot2](https://rawgit.com/eco-data-science/VisualizingData/master/ggplot2_intro.html) - a very good introduction to `ggplot2`; more information about how to deal with colors
- [Cookbook for R >> Graphs](http://www.cookbook-r.com/Graphs/) - a clear introduction to the most useful functions of `ggplot2`
- [Data Visualization with ggplot2 Cheat Sheet](https://www.rstudio.com/wp-content/uploads/2015/05/ggplot2-cheatsheet.pdf) - an official `ggplot2` cheatsheet
- [Beautiful plotting in R: A ggplot2 cheatsheet](http://zevross.com/blog/2014/08/04/beautiful-plotting-in-r-a-ggplot2-cheatsheet-3/) - a ton of examples how to customize `ggplot2` 
- [R Graph Catalog](http://shiny.stat.ubc.ca/r-graph-catalog/) - a catalog of good and bad plots
- [A Compendium of Clean Graphs in R](http://shinyapps.org/apps/RGraphCompendium/index.php) - examples of informative plots
- [Great examples of ggplot2 plots](https://rud.is/b/category/ggplot) - a lot of them...
- [Interactive visualizations with R - a minireview](http://ouzor.github.io/blog/2014/11/21/interactive-visualizations.html) - an overview of R interactive capabilities
- [plotly for R](https://cpsievert.github.io/plotly_book/) - an open online book about the `plotly` package
- [dygraphs for R](https://rstudio.github.io/dygraphs/) - an official documentation of the `dygraphs` package
- [Shiny by RStudio](http://shiny.rstudio.com/) - an official documentation of the `shiny` package